fix istio commit

Release 2.0.6-rc.2 (#1671 )
fix: Fix a typo in the README files of ai-statistics plugin (#1670 )
2026-02-25 21:21:01 +08:00 · 2025-01-15 15:29:44 +08:00 · 2025-01-14 20:10:53 +08:00 · 2025-01-14 13:39:55 +08:00 · 2025-01-14 10:03:24 +08:00 · 2025-01-14 09:48:32 +08:00
250 changed files with 9069 additions and 6168 deletions
--- a/.github/workflows/build-and-push-wasm-plugin-image.yaml
+++ b/.github/workflows/build-and-push-wasm-plugin-image.yaml
@@ -42,17 +42,19 @@ jobs:
            plugin_type="${{ github.event.inputs.plugin_type }}"
            plugin_name="${{ github.event.inputs.plugin_name }}"
            version="${{ github.event.inputs.version }}"
-            builder_image="higress-registry.cn-hangzhou.cr.aliyuncs.com/plugins/wasm-rust-builder:rust${{ env.RUST_VERSION }}-oras${{ env.ORAS_VERSION }}"
          else
            ref_name=${{ github.ref_name }}
            plugin_type=${ref_name#*-} # 删除插件类型前面的字段(wasm-)
-            plugin_type=${plugin_type%-*} # 删除插件类型后面的字段(-{plugin_name}-vX.Y.Z)
+            plugin_type=${plugin_type%%-*} # 删除插件类型后面的字段(-{plugin_name}-vX.Y.Z)
            plugin_name=${ref_name#*-*-} # 删除插件名前面的字段(wasm-go-)
            plugin_name=${plugin_name%-*} # 删除插件名后面的字段(-vX.Y.Z)
            version=$(echo "$ref_name" | awk -F'v' '{print $2}')
+          fi
+          if [[ "$plugin_type" == "rust" ]]; then
+            builder_image="higress-registry.cn-hangzhou.cr.aliyuncs.com/plugins/wasm-rust-builder:rust${{ env.RUST_VERSION }}-oras${{ env.ORAS_VERSION }}"
+          else
            builder_image="higress-registry.cn-hangzhou.cr.aliyuncs.com/plugins/wasm-go-builder:go${{ env.GO_VERSION }}-tinygo${{ env.TINYGO_VERSION }}-oras${{ env.ORAS_VERSION }}"
          fi
-
          echo "PLUGIN_TYPE=$plugin_type" >> $GITHUB_ENV
          echo "PLUGIN_NAME=$plugin_name" >> $GITHUB_ENV
          echo "VERSION=$version" >> $GITHUB_ENV
--- a/.github/workflows/helm-docs.yaml
+++ b/.github/workflows/helm-docs.yaml
@@ -0,0 +1,35 @@
+name: "Helm Docs"
+
+on:
+  pull_request:
+    branches:
+      - "*"
+
+  push:
+
+jobs:
+
+  helm:
+    name: Helm Docs
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v4
+        with:
+          fetch-depth: 1
+
+      - name: Setup Go
+        uses: actions/setup-go@v5
+        with:
+          go-version: '1.22.9'
+
+      - name: Run helm-docs
+        run: |
+          GOBIN=$PWD GO111MODULE=on go install github.com/norwoodj/helm-docs/cmd/helm-docs@v1.14.2
+          ./helm-docs -c ${GITHUB_WORKSPACE}/helm/higress -f ../core/values.yaml
+          DIFF=$(git diff ${GITHUB_WORKSPACE}/helm/higress/*md)
+          if [ ! -z "$DIFF" ]; then
+            echo "Please use helm-docs in your clone, of your fork, of the project, and commit a updated README.md for the chart."
+          fi
+          git diff --exit-code
+          rm -f ./helm-docs
--- a/.github/workflows/release-hgctl.yaml
+++ b/.github/workflows/release-hgctl.yaml
@@ -58,7 +58,7 @@ jobs:
          hgctl_${{ env.HGCTL_VERSION }}_darwin_arm64.tar.gz

  release-hgctl-macos-amd64:
-    runs-on: macos-12
+    runs-on: macos-14
    env:
      HGCTL_VERSION: ${{github.ref_name}}
    steps:
--- a/.gitignore
+++ b/.gitignore
@@ -16,4 +16,5 @@ helm/**/charts/**.tgz
 target/
 tools/hack/cluster.conf
 envoy/1.20
-istio/1.12
+istio/1.12
+Cargo.lock
--- a/.licenserc.yaml
+++ b/.licenserc.yaml
@@ -12,6 +12,7 @@ header:
    - 'LICENSE'
    - 'api/**'
    - 'samples/**'
+    - 'docs/**'
    - '.github/**'
    - '.licenserc.yaml'
    - 'helm/**'
--- a/3
+++ b/3
@@ -2,7 +2,8 @@
 /envoy @gengleilei @johnlanni
 /istio @SpecialYang @johnlanni
 /pkg @SpecialYang @johnlanni @CH3CHO
-/plugins @johnlanni @WeixinX @CH3CHO
+/plugins @johnlanni @CH3CHO @rinfx
+/plugins/wasm-go/extensions/ai-proxy @cr7258 @CH3CHO @rinfx
 /plugins/wasm-rust @007gzs @jizhuozhi
 /registry @NameHaibinZhang @2456868764 @johnlanni
 /test @Xunzhuo @2456868764 @CH3CHO
--- a/Makefile.core.mk
+++ b/Makefile.core.mk
@@ -144,7 +144,7 @@ docker-buildx-push: clean-env docker.higress-buildx
 export PARENT_GIT_TAG:=$(shell cat VERSION)
 export PARENT_GIT_REVISION:=$(TAG)

-export ENVOY_PACKAGE_URL_PATTERN?=https://github.com/higress-group/proxy/releases/download/v2.0.0/envoy-symbol-ARCH.tar.gz
+export ENVOY_PACKAGE_URL_PATTERN?=https://github.com/higress-group/proxy/releases/download/v2.1.0/envoy-symbol-ARCH.tar.gz

 build-envoy: prebuild
 	./tools/hack/build-envoy.sh
@@ -187,8 +187,8 @@ install: pre-install
 	cd helm/higress; helm dependency build
 	helm install higress helm/higress -n higress-system --create-namespace --set 'global.local=true'

-ENVOY_LATEST_IMAGE_TAG ?= 2.0.1
-ISTIO_LATEST_IMAGE_TAG ?= 2.0.1
+ENVOY_LATEST_IMAGE_TAG ?= 958467a353d411ae3f06e03b096bfd342cddb2c6
+ISTIO_LATEST_IMAGE_TAG ?= 01ad224eff2bb7eb200869fc64221f739a48e07e

 install-dev: pre-install
 	helm install higress helm/core -n higress-system --create-namespace --set 'controller.tag=$(TAG)' --set 'gateway.replicas=1' --set 'pilot.tag=$(ISTIO_LATEST_IMAGE_TAG)' --set 'gateway.tag=$(ENVOY_LATEST_IMAGE_TAG)' --set 'global.local=true'
@@ -299,7 +299,7 @@ kube-load-image: $(tools/kind) ## Install the Higress image to a kind cluster us
 	tools/hack/docker-pull-image.sh higress-registry.cn-hangzhou.cr.aliyuncs.com/higress/echo-server 1.3.0
 	tools/hack/docker-pull-image.sh higress-registry.cn-hangzhou.cr.aliyuncs.com/higress/echo-server v1.0
 	tools/hack/docker-pull-image.sh higress-registry.cn-hangzhou.cr.aliyuncs.com/higress/echo-body 1.0.0
-	tools/hack/docker-pull-image.sh openpolicyagent/opa latest
+	tools/hack/docker-pull-image.sh openpolicyagent/opa 0.61.0
 	tools/hack/docker-pull-image.sh curlimages/curl latest
 	tools/hack/docker-pull-image.sh registry.cn-hangzhou.aliyuncs.com/2456868764/httpbin 1.0.2
 	tools/hack/docker-pull-image.sh registry.cn-hangzhou.aliyuncs.com/hinsteny/nacos-standlone-rc3 1.0.0-RC3
@@ -312,7 +312,7 @@ kube-load-image: $(tools/kind) ## Install the Higress image to a kind cluster us
 	tools/hack/kind-load-image.sh higress-registry.cn-hangzhou.cr.aliyuncs.com/higress/echo-server 1.3.0
 	tools/hack/kind-load-image.sh higress-registry.cn-hangzhou.cr.aliyuncs.com/higress/echo-server v1.0
 	tools/hack/kind-load-image.sh higress-registry.cn-hangzhou.cr.aliyuncs.com/higress/echo-body 1.0.0
-	tools/hack/kind-load-image.sh openpolicyagent/opa latest
+	tools/hack/kind-load-image.sh openpolicyagent/opa 0.61.0
 	tools/hack/kind-load-image.sh curlimages/curl latest
 	tools/hack/kind-load-image.sh registry.cn-hangzhou.aliyuncs.com/2456868764/httpbin 1.0.2
 	tools/hack/kind-load-image.sh registry.cn-hangzhou.aliyuncs.com/hinsteny/nacos-standlone-rc3 1.0.0-RC3
--- a/README.md
+++ b/README.md
@@ -6,9 +6,14 @@
 </h1>
 <h4 align="center"> AI Native API Gateway </h4>

+<div align="center">
+    
 [![Build Status](https://github.com/alibaba/higress/actions/workflows/build-and-test.yaml/badge.svg?branch=main)](https://github.com/alibaba/higress/actions)
 [![license](https://img.shields.io/github/license/alibaba/higress.svg)](https://www.apache.org/licenses/LICENSE-2.0.html)

+<a href="https://trendshift.io/repositories/10918" target="_blank"><img src="https://trendshift.io/api/badge/repositories/10918" alt="alibaba%2Fhigress | Trendshift" style="width: 250px; height: 55px;" width="250" height="55"/></a>
+</div>
+
 [**官网**](https://higress.cn/) &nbsp; |
 &nbsp; [**文档**](https://higress.cn/docs/latest/overview/what-is-higress/) &nbsp; |
 &nbsp; [**博客**](https://higress.cn/blog/) &nbsp; |
@@ -17,6 +22,7 @@
 &nbsp; [**AI插件**](https://higress.cn/plugin/) &nbsp;


+
 <p>
   <a href="README_EN.md"> English <a/>| 中文 | <a href="README_JP.md"> 日本語 <a/> 
 </p>
@@ -64,6 +70,10 @@ docker run -d --rm --name higress-ai -v ${PWD}:/data \

 K8s 下使用 Helm 部署等其他安装方式可以参考官网 [Quick Start 文档](https://higress.cn/docs/latest/user/quickstart/)。

+如果您是在云上部署，生产环境推荐使用[企业版](https://higress.io/cloud/)，开发测试可以使用下面一键部署社区版：
+
+[![Deploy on AlibabaCloud ComputeNest](https://service-info-public.oss-cn-hangzhou.aliyuncs.com/computenest.svg)](https://computenest.console.aliyun.com/service/instance/create/default?type=user&ServiceName=Higress社区版)
+

 ## 使用场景

@@ -176,7 +186,7 @@ K8s 下使用 Helm 部署等其他安装方式可以参考官网 [Quick Start

 ### 交流群

-![image](https://img.alicdn.com/imgextra/i2/O1CN01BkopaB22ZsvamFftE_!!6000000007135-0-tps-720-405.jpg)
+![image](https://img.alicdn.com/imgextra/i2/O1CN01fZefEP1aPWkzG3A19_!!6000000003322-0-tps-720-405.jpg)

 ### 技术分享

--- a/2
+++ b/2
@@ -1 +1 @@
-v2.0.3
+v2.0.6-rc.2
--- a/api/extensions/v1alpha1/wasmplugin.pb.go
+++ b/api/extensions/v1alpha1/wasmplugin.pb.go
@@ -341,7 +341,7 @@ type WasmPlugin struct {
 	// Extended by Higress, matching rules take effect
 	MatchRules []*MatchRule `protobuf:"bytes,102,rep,name=match_rules,json=matchRules,proto3" json:"match_rules,omitempty"`
 	// disable the default config
-	DefaultConfigDisable bool `protobuf:"varint,103,opt,name=default_config_disable,json=defaultConfigDisable,proto3" json:"default_config_disable,omitempty"`
+	DefaultConfigDisable *wrappers.BoolValue `protobuf:"bytes,103,opt,name=default_config_disable,json=defaultConfigDisable,proto3" json:"default_config_disable,omitempty"`
 }

 func (x *WasmPlugin) Reset() {
@@ -467,11 +467,11 @@ func (x *WasmPlugin) GetMatchRules() []*MatchRule {
 	return nil
 }

-func (x *WasmPlugin) GetDefaultConfigDisable() bool {
+func (x *WasmPlugin) GetDefaultConfigDisable() *wrappers.BoolValue {
 	if x != nil {
 		return x.DefaultConfigDisable
 	}
-	return false
+	return nil
 }

 // Extended by Higress
@@ -480,11 +480,11 @@ type MatchRule struct {
 	sizeCache     protoimpl.SizeCache
 	unknownFields protoimpl.UnknownFields

-	Ingress       []string        `protobuf:"bytes,1,rep,name=ingress,proto3" json:"ingress,omitempty"`
-	Domain        []string        `protobuf:"bytes,2,rep,name=domain,proto3" json:"domain,omitempty"`
-	Config        *_struct.Struct `protobuf:"bytes,3,opt,name=config,proto3" json:"config,omitempty"`
-	ConfigDisable bool            `protobuf:"varint,4,opt,name=config_disable,json=configDisable,proto3" json:"config_disable,omitempty"`
-	Service       []string        `protobuf:"bytes,5,rep,name=service,proto3" json:"service,omitempty"`
+	Ingress       []string            `protobuf:"bytes,1,rep,name=ingress,proto3" json:"ingress,omitempty"`
+	Domain        []string            `protobuf:"bytes,2,rep,name=domain,proto3" json:"domain,omitempty"`
+	Config        *_struct.Struct     `protobuf:"bytes,3,opt,name=config,proto3" json:"config,omitempty"`
+	ConfigDisable *wrappers.BoolValue `protobuf:"bytes,4,opt,name=config_disable,json=configDisable,proto3" json:"config_disable,omitempty"`
+	Service       []string            `protobuf:"bytes,5,rep,name=service,proto3" json:"service,omitempty"`
 }

 func (x *MatchRule) Reset() {
@@ -540,11 +540,11 @@ func (x *MatchRule) GetConfig() *_struct.Struct {
 	return nil
 }

-func (x *MatchRule) GetConfigDisable() bool {
+func (x *MatchRule) GetConfigDisable() *wrappers.BoolValue {
 	if x != nil {
 		return x.ConfigDisable
 	}
-	return false
+	return nil
 }

 func (x *MatchRule) GetService() []string {
@@ -686,7 +686,7 @@ var file_extensions_v1alpha1_wasmplugin_proto_rawDesc = []byte{
 	0x6f, 0x62, 0x75, 0x66, 0x2f, 0x77, 0x72, 0x61, 0x70, 0x70, 0x65, 0x72, 0x73, 0x2e, 0x70, 0x72,
 	0x6f, 0x74, 0x6f, 0x1a, 0x1c, 0x67, 0x6f, 0x6f, 0x67, 0x6c, 0x65, 0x2f, 0x70, 0x72, 0x6f, 0x74,
 	0x6f, 0x62, 0x75, 0x66, 0x2f, 0x73, 0x74, 0x72, 0x75, 0x63, 0x74, 0x2e, 0x70, 0x72, 0x6f, 0x74,
-	0x6f, 0x22, 0x8d, 0x06, 0x0a, 0x0a, 0x57, 0x61, 0x73, 0x6d, 0x50, 0x6c, 0x75, 0x67, 0x69, 0x6e,
+	0x6f, 0x22, 0xa9, 0x06, 0x0a, 0x0a, 0x57, 0x61, 0x73, 0x6d, 0x50, 0x6c, 0x75, 0x67, 0x69, 0x6e,
 	0x12, 0x10, 0x0a, 0x03, 0x75, 0x72, 0x6c, 0x18, 0x02, 0x20, 0x01, 0x28, 0x09, 0x52, 0x03, 0x75,
 	0x72, 0x6c, 0x12, 0x16, 0x0a, 0x06, 0x73, 0x68, 0x61, 0x32, 0x35, 0x36, 0x18, 0x03, 0x20, 0x01,
 	0x28, 0x09, 0x52, 0x06, 0x73, 0x68, 0x61, 0x32, 0x35, 0x36, 0x12, 0x53, 0x0a, 0x11, 0x69, 0x6d,
@@ -731,52 +731,55 @@ var file_extensions_v1alpha1_wasmplugin_proto_rawDesc = []byte{
 	0x73, 0x18, 0x66, 0x20, 0x03, 0x28, 0x0b, 0x32, 0x26, 0x2e, 0x68, 0x69, 0x67, 0x72, 0x65, 0x73,
 	0x73, 0x2e, 0x65, 0x78, 0x74, 0x65, 0x6e, 0x73, 0x69, 0x6f, 0x6e, 0x73, 0x2e, 0x76, 0x31, 0x61,
 	0x6c, 0x70, 0x68, 0x61, 0x31, 0x2e, 0x4d, 0x61, 0x74, 0x63, 0x68, 0x52, 0x75, 0x6c, 0x65, 0x52,
-	0x0a, 0x6d, 0x61, 0x74, 0x63, 0x68, 0x52, 0x75, 0x6c, 0x65, 0x73, 0x12, 0x34, 0x0a, 0x16, 0x64,
+	0x0a, 0x6d, 0x61, 0x74, 0x63, 0x68, 0x52, 0x75, 0x6c, 0x65, 0x73, 0x12, 0x50, 0x0a, 0x16, 0x64,
 	0x65, 0x66, 0x61, 0x75, 0x6c, 0x74, 0x5f, 0x63, 0x6f, 0x6e, 0x66, 0x69, 0x67, 0x5f, 0x64, 0x69,
-	0x73, 0x61, 0x62, 0x6c, 0x65, 0x18, 0x67, 0x20, 0x01, 0x28, 0x08, 0x52, 0x14, 0x64, 0x65, 0x66,
-	0x61, 0x75, 0x6c, 0x74, 0x43, 0x6f, 0x6e, 0x66, 0x69, 0x67, 0x44, 0x69, 0x73, 0x61, 0x62, 0x6c,
-	0x65, 0x22, 0xaf, 0x01, 0x0a, 0x09, 0x4d, 0x61, 0x74, 0x63, 0x68, 0x52, 0x75, 0x6c, 0x65, 0x12,
-	0x18, 0x0a, 0x07, 0x69, 0x6e, 0x67, 0x72, 0x65, 0x73, 0x73, 0x18, 0x01, 0x20, 0x03, 0x28, 0x09,
-	0x52, 0x07, 0x69, 0x6e, 0x67, 0x72, 0x65, 0x73, 0x73, 0x12, 0x16, 0x0a, 0x06, 0x64, 0x6f, 0x6d,
-	0x61, 0x69, 0x6e, 0x18, 0x02, 0x20, 0x03, 0x28, 0x09, 0x52, 0x06, 0x64, 0x6f, 0x6d, 0x61, 0x69,
-	0x6e, 0x12, 0x2f, 0x0a, 0x06, 0x63, 0x6f, 0x6e, 0x66, 0x69, 0x67, 0x18, 0x03, 0x20, 0x01, 0x28,
-	0x0b, 0x32, 0x17, 0x2e, 0x67, 0x6f, 0x6f, 0x67, 0x6c, 0x65, 0x2e, 0x70, 0x72, 0x6f, 0x74, 0x6f,
-	0x62, 0x75, 0x66, 0x2e, 0x53, 0x74, 0x72, 0x75, 0x63, 0x74, 0x52, 0x06, 0x63, 0x6f, 0x6e, 0x66,
-	0x69, 0x67, 0x12, 0x25, 0x0a, 0x0e, 0x63, 0x6f, 0x6e, 0x66, 0x69, 0x67, 0x5f, 0x64, 0x69, 0x73,
-	0x61, 0x62, 0x6c, 0x65, 0x18, 0x04, 0x20, 0x01, 0x28, 0x08, 0x52, 0x0d, 0x63, 0x6f, 0x6e, 0x66,
-	0x69, 0x67, 0x44, 0x69, 0x73, 0x61, 0x62, 0x6c, 0x65, 0x12, 0x18, 0x0a, 0x07, 0x73, 0x65, 0x72,
-	0x76, 0x69, 0x63, 0x65, 0x18, 0x05, 0x20, 0x03, 0x28, 0x09, 0x52, 0x07, 0x73, 0x65, 0x72, 0x76,
-	0x69, 0x63, 0x65, 0x22, 0x41, 0x0a, 0x08, 0x56, 0x6d, 0x43, 0x6f, 0x6e, 0x66, 0x69, 0x67, 0x12,
-	0x35, 0x0a, 0x03, 0x65, 0x6e, 0x76, 0x18, 0x01, 0x20, 0x03, 0x28, 0x0b, 0x32, 0x23, 0x2e, 0x68,
-	0x69, 0x67, 0x72, 0x65, 0x73, 0x73, 0x2e, 0x65, 0x78, 0x74, 0x65, 0x6e, 0x73, 0x69, 0x6f, 0x6e,
-	0x73, 0x2e, 0x76, 0x31, 0x61, 0x6c, 0x70, 0x68, 0x61, 0x31, 0x2e, 0x45, 0x6e, 0x76, 0x56, 0x61,
-	0x72, 0x52, 0x03, 0x65, 0x6e, 0x76, 0x22, 0x7e, 0x0a, 0x06, 0x45, 0x6e, 0x76, 0x56, 0x61, 0x72,
-	0x12, 0x12, 0x0a, 0x04, 0x6e, 0x61, 0x6d, 0x65, 0x18, 0x01, 0x20, 0x01, 0x28, 0x09, 0x52, 0x04,
-	0x6e, 0x61, 0x6d, 0x65, 0x12, 0x4a, 0x0a, 0x0a, 0x76, 0x61, 0x6c, 0x75, 0x65, 0x5f, 0x66, 0x72,
-	0x6f, 0x6d, 0x18, 0x03, 0x20, 0x01, 0x28, 0x0e, 0x32, 0x2b, 0x2e, 0x68, 0x69, 0x67, 0x72, 0x65,
-	0x73, 0x73, 0x2e, 0x65, 0x78, 0x74, 0x65, 0x6e, 0x73, 0x69, 0x6f, 0x6e, 0x73, 0x2e, 0x76, 0x31,
-	0x61, 0x6c, 0x70, 0x68, 0x61, 0x31, 0x2e, 0x45, 0x6e, 0x76, 0x56, 0x61, 0x6c, 0x75, 0x65, 0x53,
-	0x6f, 0x75, 0x72, 0x63, 0x65, 0x52, 0x09, 0x76, 0x61, 0x6c, 0x75, 0x65, 0x46, 0x72, 0x6f, 0x6d,
-	0x12, 0x14, 0x0a, 0x05, 0x76, 0x61, 0x6c, 0x75, 0x65, 0x18, 0x02, 0x20, 0x01, 0x28, 0x09, 0x52,
-	0x05, 0x76, 0x61, 0x6c, 0x75, 0x65, 0x2a, 0x45, 0x0a, 0x0b, 0x50, 0x6c, 0x75, 0x67, 0x69, 0x6e,
-	0x50, 0x68, 0x61, 0x73, 0x65, 0x12, 0x15, 0x0a, 0x11, 0x55, 0x4e, 0x53, 0x50, 0x45, 0x43, 0x49,
-	0x46, 0x49, 0x45, 0x44, 0x5f, 0x50, 0x48, 0x41, 0x53, 0x45, 0x10, 0x00, 0x12, 0x09, 0x0a, 0x05,
-	0x41, 0x55, 0x54, 0x48, 0x4e, 0x10, 0x01, 0x12, 0x09, 0x0a, 0x05, 0x41, 0x55, 0x54, 0x48, 0x5a,
-	0x10, 0x02, 0x12, 0x09, 0x0a, 0x05, 0x53, 0x54, 0x41, 0x54, 0x53, 0x10, 0x03, 0x2a, 0x42, 0x0a,
-	0x0a, 0x50, 0x75, 0x6c, 0x6c, 0x50, 0x6f, 0x6c, 0x69, 0x63, 0x79, 0x12, 0x16, 0x0a, 0x12, 0x55,
-	0x4e, 0x53, 0x50, 0x45, 0x43, 0x49, 0x46, 0x49, 0x45, 0x44, 0x5f, 0x50, 0x4f, 0x4c, 0x49, 0x43,
-	0x59, 0x10, 0x00, 0x12, 0x10, 0x0a, 0x0c, 0x49, 0x66, 0x4e, 0x6f, 0x74, 0x50, 0x72, 0x65, 0x73,
-	0x65, 0x6e, 0x74, 0x10, 0x01, 0x12, 0x0a, 0x0a, 0x06, 0x41, 0x6c, 0x77, 0x61, 0x79, 0x73, 0x10,
-	0x02, 0x2a, 0x26, 0x0a, 0x0e, 0x45, 0x6e, 0x76, 0x56, 0x61, 0x6c, 0x75, 0x65, 0x53, 0x6f, 0x75,
-	0x72, 0x63, 0x65, 0x12, 0x0a, 0x0a, 0x06, 0x49, 0x4e, 0x4c, 0x49, 0x4e, 0x45, 0x10, 0x00, 0x12,
-	0x08, 0x0a, 0x04, 0x48, 0x4f, 0x53, 0x54, 0x10, 0x01, 0x2a, 0x2d, 0x0a, 0x0c, 0x46, 0x61, 0x69,
-	0x6c, 0x53, 0x74, 0x72, 0x61, 0x74, 0x65, 0x67, 0x79, 0x12, 0x0e, 0x0a, 0x0a, 0x46, 0x41, 0x49,
-	0x4c, 0x5f, 0x43, 0x4c, 0x4f, 0x53, 0x45, 0x10, 0x00, 0x12, 0x0d, 0x0a, 0x09, 0x46, 0x41, 0x49,
-	0x4c, 0x5f, 0x4f, 0x50, 0x45, 0x4e, 0x10, 0x01, 0x42, 0x34, 0x5a, 0x32, 0x67, 0x69, 0x74, 0x68,
-	0x75, 0x62, 0x2e, 0x63, 0x6f, 0x6d, 0x2f, 0x61, 0x6c, 0x69, 0x62, 0x61, 0x62, 0x61, 0x2f, 0x68,
-	0x69, 0x67, 0x72, 0x65, 0x73, 0x73, 0x2f, 0x61, 0x70, 0x69, 0x2f, 0x65, 0x78, 0x74, 0x65, 0x6e,
-	0x73, 0x69, 0x6f, 0x6e, 0x73, 0x2f, 0x76, 0x31, 0x61, 0x6c, 0x70, 0x68, 0x61, 0x31, 0x62, 0x06,
-	0x70, 0x72, 0x6f, 0x74, 0x6f, 0x33,
+	0x73, 0x61, 0x62, 0x6c, 0x65, 0x18, 0x67, 0x20, 0x01, 0x28, 0x0b, 0x32, 0x1a, 0x2e, 0x67, 0x6f,
+	0x6f, 0x67, 0x6c, 0x65, 0x2e, 0x70, 0x72, 0x6f, 0x74, 0x6f, 0x62, 0x75, 0x66, 0x2e, 0x42, 0x6f,
+	0x6f, 0x6c, 0x56, 0x61, 0x6c, 0x75, 0x65, 0x52, 0x14, 0x64, 0x65, 0x66, 0x61, 0x75, 0x6c, 0x74,
+	0x43, 0x6f, 0x6e, 0x66, 0x69, 0x67, 0x44, 0x69, 0x73, 0x61, 0x62, 0x6c, 0x65, 0x22, 0xcb, 0x01,
+	0x0a, 0x09, 0x4d, 0x61, 0x74, 0x63, 0x68, 0x52, 0x75, 0x6c, 0x65, 0x12, 0x18, 0x0a, 0x07, 0x69,
+	0x6e, 0x67, 0x72, 0x65, 0x73, 0x73, 0x18, 0x01, 0x20, 0x03, 0x28, 0x09, 0x52, 0x07, 0x69, 0x6e,
+	0x67, 0x72, 0x65, 0x73, 0x73, 0x12, 0x16, 0x0a, 0x06, 0x64, 0x6f, 0x6d, 0x61, 0x69, 0x6e, 0x18,
+	0x02, 0x20, 0x03, 0x28, 0x09, 0x52, 0x06, 0x64, 0x6f, 0x6d, 0x61, 0x69, 0x6e, 0x12, 0x2f, 0x0a,
+	0x06, 0x63, 0x6f, 0x6e, 0x66, 0x69, 0x67, 0x18, 0x03, 0x20, 0x01, 0x28, 0x0b, 0x32, 0x17, 0x2e,
+	0x67, 0x6f, 0x6f, 0x67, 0x6c, 0x65, 0x2e, 0x70, 0x72, 0x6f, 0x74, 0x6f, 0x62, 0x75, 0x66, 0x2e,
+	0x53, 0x74, 0x72, 0x75, 0x63, 0x74, 0x52, 0x06, 0x63, 0x6f, 0x6e, 0x66, 0x69, 0x67, 0x12, 0x41,
+	0x0a, 0x0e, 0x63, 0x6f, 0x6e, 0x66, 0x69, 0x67, 0x5f, 0x64, 0x69, 0x73, 0x61, 0x62, 0x6c, 0x65,
+	0x18, 0x04, 0x20, 0x01, 0x28, 0x0b, 0x32, 0x1a, 0x2e, 0x67, 0x6f, 0x6f, 0x67, 0x6c, 0x65, 0x2e,
+	0x70, 0x72, 0x6f, 0x74, 0x6f, 0x62, 0x75, 0x66, 0x2e, 0x42, 0x6f, 0x6f, 0x6c, 0x56, 0x61, 0x6c,
+	0x75, 0x65, 0x52, 0x0d, 0x63, 0x6f, 0x6e, 0x66, 0x69, 0x67, 0x44, 0x69, 0x73, 0x61, 0x62, 0x6c,
+	0x65, 0x12, 0x18, 0x0a, 0x07, 0x73, 0x65, 0x72, 0x76, 0x69, 0x63, 0x65, 0x18, 0x05, 0x20, 0x03,
+	0x28, 0x09, 0x52, 0x07, 0x73, 0x65, 0x72, 0x76, 0x69, 0x63, 0x65, 0x22, 0x41, 0x0a, 0x08, 0x56,
+	0x6d, 0x43, 0x6f, 0x6e, 0x66, 0x69, 0x67, 0x12, 0x35, 0x0a, 0x03, 0x65, 0x6e, 0x76, 0x18, 0x01,
+	0x20, 0x03, 0x28, 0x0b, 0x32, 0x23, 0x2e, 0x68, 0x69, 0x67, 0x72, 0x65, 0x73, 0x73, 0x2e, 0x65,
+	0x78, 0x74, 0x65, 0x6e, 0x73, 0x69, 0x6f, 0x6e, 0x73, 0x2e, 0x76, 0x31, 0x61, 0x6c, 0x70, 0x68,
+	0x61, 0x31, 0x2e, 0x45, 0x6e, 0x76, 0x56, 0x61, 0x72, 0x52, 0x03, 0x65, 0x6e, 0x76, 0x22, 0x7e,
+	0x0a, 0x06, 0x45, 0x6e, 0x76, 0x56, 0x61, 0x72, 0x12, 0x12, 0x0a, 0x04, 0x6e, 0x61, 0x6d, 0x65,
+	0x18, 0x01, 0x20, 0x01, 0x28, 0x09, 0x52, 0x04, 0x6e, 0x61, 0x6d, 0x65, 0x12, 0x4a, 0x0a, 0x0a,
+	0x76, 0x61, 0x6c, 0x75, 0x65, 0x5f, 0x66, 0x72, 0x6f, 0x6d, 0x18, 0x03, 0x20, 0x01, 0x28, 0x0e,
+	0x32, 0x2b, 0x2e, 0x68, 0x69, 0x67, 0x72, 0x65, 0x73, 0x73, 0x2e, 0x65, 0x78, 0x74, 0x65, 0x6e,
+	0x73, 0x69, 0x6f, 0x6e, 0x73, 0x2e, 0x76, 0x31, 0x61, 0x6c, 0x70, 0x68, 0x61, 0x31, 0x2e, 0x45,
+	0x6e, 0x76, 0x56, 0x61, 0x6c, 0x75, 0x65, 0x53, 0x6f, 0x75, 0x72, 0x63, 0x65, 0x52, 0x09, 0x76,
+	0x61, 0x6c, 0x75, 0x65, 0x46, 0x72, 0x6f, 0x6d, 0x12, 0x14, 0x0a, 0x05, 0x76, 0x61, 0x6c, 0x75,
+	0x65, 0x18, 0x02, 0x20, 0x01, 0x28, 0x09, 0x52, 0x05, 0x76, 0x61, 0x6c, 0x75, 0x65, 0x2a, 0x45,
+	0x0a, 0x0b, 0x50, 0x6c, 0x75, 0x67, 0x69, 0x6e, 0x50, 0x68, 0x61, 0x73, 0x65, 0x12, 0x15, 0x0a,
+	0x11, 0x55, 0x4e, 0x53, 0x50, 0x45, 0x43, 0x49, 0x46, 0x49, 0x45, 0x44, 0x5f, 0x50, 0x48, 0x41,
+	0x53, 0x45, 0x10, 0x00, 0x12, 0x09, 0x0a, 0x05, 0x41, 0x55, 0x54, 0x48, 0x4e, 0x10, 0x01, 0x12,
+	0x09, 0x0a, 0x05, 0x41, 0x55, 0x54, 0x48, 0x5a, 0x10, 0x02, 0x12, 0x09, 0x0a, 0x05, 0x53, 0x54,
+	0x41, 0x54, 0x53, 0x10, 0x03, 0x2a, 0x42, 0x0a, 0x0a, 0x50, 0x75, 0x6c, 0x6c, 0x50, 0x6f, 0x6c,
+	0x69, 0x63, 0x79, 0x12, 0x16, 0x0a, 0x12, 0x55, 0x4e, 0x53, 0x50, 0x45, 0x43, 0x49, 0x46, 0x49,
+	0x45, 0x44, 0x5f, 0x50, 0x4f, 0x4c, 0x49, 0x43, 0x59, 0x10, 0x00, 0x12, 0x10, 0x0a, 0x0c, 0x49,
+	0x66, 0x4e, 0x6f, 0x74, 0x50, 0x72, 0x65, 0x73, 0x65, 0x6e, 0x74, 0x10, 0x01, 0x12, 0x0a, 0x0a,
+	0x06, 0x41, 0x6c, 0x77, 0x61, 0x79, 0x73, 0x10, 0x02, 0x2a, 0x26, 0x0a, 0x0e, 0x45, 0x6e, 0x76,
+	0x56, 0x61, 0x6c, 0x75, 0x65, 0x53, 0x6f, 0x75, 0x72, 0x63, 0x65, 0x12, 0x0a, 0x0a, 0x06, 0x49,
+	0x4e, 0x4c, 0x49, 0x4e, 0x45, 0x10, 0x00, 0x12, 0x08, 0x0a, 0x04, 0x48, 0x4f, 0x53, 0x54, 0x10,
+	0x01, 0x2a, 0x2d, 0x0a, 0x0c, 0x46, 0x61, 0x69, 0x6c, 0x53, 0x74, 0x72, 0x61, 0x74, 0x65, 0x67,
+	0x79, 0x12, 0x0e, 0x0a, 0x0a, 0x46, 0x41, 0x49, 0x4c, 0x5f, 0x43, 0x4c, 0x4f, 0x53, 0x45, 0x10,
+	0x00, 0x12, 0x0d, 0x0a, 0x09, 0x46, 0x41, 0x49, 0x4c, 0x5f, 0x4f, 0x50, 0x45, 0x4e, 0x10, 0x01,
+	0x42, 0x34, 0x5a, 0x32, 0x67, 0x69, 0x74, 0x68, 0x75, 0x62, 0x2e, 0x63, 0x6f, 0x6d, 0x2f, 0x61,
+	0x6c, 0x69, 0x62, 0x61, 0x62, 0x61, 0x2f, 0x68, 0x69, 0x67, 0x72, 0x65, 0x73, 0x73, 0x2f, 0x61,
+	0x70, 0x69, 0x2f, 0x65, 0x78, 0x74, 0x65, 0x6e, 0x73, 0x69, 0x6f, 0x6e, 0x73, 0x2f, 0x76, 0x31,
+	0x61, 0x6c, 0x70, 0x68, 0x61, 0x31, 0x62, 0x06, 0x70, 0x72, 0x6f, 0x74, 0x6f, 0x33,
 }

 var (
@@ -804,6 +807,7 @@ var file_extensions_v1alpha1_wasmplugin_proto_goTypes = []interface{}{
 	(*EnvVar)(nil),              // 7: higress.extensions.v1alpha1.EnvVar
 	(*_struct.Struct)(nil),      // 8: google.protobuf.Struct
 	(*wrappers.Int32Value)(nil), // 9: google.protobuf.Int32Value
+	(*wrappers.BoolValue)(nil),  // 10: google.protobuf.BoolValue
 }
 var file_extensions_v1alpha1_wasmplugin_proto_depIdxs = []int32{
 	1,  // 0: higress.extensions.v1alpha1.WasmPlugin.image_pull_policy:type_name -> higress.extensions.v1alpha1.PullPolicy
@@ -814,14 +818,16 @@ var file_extensions_v1alpha1_wasmplugin_proto_depIdxs = []int32{
 	6,  // 5: higress.extensions.v1alpha1.WasmPlugin.vm_config:type_name -> higress.extensions.v1alpha1.VmConfig
 	8,  // 6: higress.extensions.v1alpha1.WasmPlugin.default_config:type_name -> google.protobuf.Struct
 	5,  // 7: higress.extensions.v1alpha1.WasmPlugin.match_rules:type_name -> higress.extensions.v1alpha1.MatchRule
-	8,  // 8: higress.extensions.v1alpha1.MatchRule.config:type_name -> google.protobuf.Struct
-	7,  // 9: higress.extensions.v1alpha1.VmConfig.env:type_name -> higress.extensions.v1alpha1.EnvVar
-	2,  // 10: higress.extensions.v1alpha1.EnvVar.value_from:type_name -> higress.extensions.v1alpha1.EnvValueSource
-	11, // [11:11] is the sub-list for method output_type
-	11, // [11:11] is the sub-list for method input_type
-	11, // [11:11] is the sub-list for extension type_name
-	11, // [11:11] is the sub-list for extension extendee
-	0,  // [0:11] is the sub-list for field type_name
+	10, // 8: higress.extensions.v1alpha1.WasmPlugin.default_config_disable:type_name -> google.protobuf.BoolValue
+	8,  // 9: higress.extensions.v1alpha1.MatchRule.config:type_name -> google.protobuf.Struct
+	10, // 10: higress.extensions.v1alpha1.MatchRule.config_disable:type_name -> google.protobuf.BoolValue
+	7,  // 11: higress.extensions.v1alpha1.VmConfig.env:type_name -> higress.extensions.v1alpha1.EnvVar
+	2,  // 12: higress.extensions.v1alpha1.EnvVar.value_from:type_name -> higress.extensions.v1alpha1.EnvValueSource
+	13, // [13:13] is the sub-list for method output_type
+	13, // [13:13] is the sub-list for method input_type
+	13, // [13:13] is the sub-list for extension type_name
+	13, // [13:13] is the sub-list for extension extendee
+	0,  // [0:13] is the sub-list for field type_name
 }

 func init() { file_extensions_v1alpha1_wasmplugin_proto_init() }
--- a/api/extensions/v1alpha1/wasmplugin.proto
+++ b/api/extensions/v1alpha1/wasmplugin.proto
@@ -112,7 +112,7 @@ message WasmPlugin {
  // Extended by Higress, matching rules take effect
  repeated MatchRule match_rules = 102;
  // disable the default config
-  bool default_config_disable = 103;
+  google.protobuf.BoolValue default_config_disable = 103;
 }

 // Extended by Higress
@@ -120,7 +120,7 @@ message MatchRule {
  repeated string ingress = 1;
  repeated string domain = 2;
  google.protobuf.Struct config = 3;
-  bool config_disable = 4;
+  google.protobuf.BoolValue config_disable = 4;
  repeated string service = 5;
 }

--- a/docs/architecture.md
+++ b/docs/architecture.md
@@ -0,0 +1,143 @@
+# Higress 核心组件和原理
+
+Higress 是基于 Envoy 和 Istio 进行二次定制化开发构建和功能增强，同时利用 Envoy 和 Istio 一些插件机制，实现了一个轻量级的网关服务。其包括 3 个核心组件：Higress Controller（控制器）、Higress Gateway（网关）和 Higress Console（控制台）。
+下图概况了其核心工作流程：
+
+![img](./images/img_02_01.png)
+
+本章将重点介绍 Higress 的两个核心组件：Higress Controller 和 Higress Gateway。
+
+## 1 Higress Console
+
+Higress Console 是 Higress 网关的管理控制台，主要功能是管理 Higress 网关的路由配置、插件配置等。
+
+### 1.1 Higress Admin SDK
+
+Higress Admin SDK 脱胎于 Higress Console。起初，它作为 Higress Console 的一部分，为前端界面提供实际的功能支持。后来考虑到对接外部系统等需求，将配置管理的部分剥离出来，形成一个独立的逻辑组件，便于和各个系统进行对接。目前支持服务来源管理、服务管理、路由管理、域名管理、证书管理、插件管理等功能。
+Higress Admin SDK 现在只提供 Java 版本，且要求 JDK 版本不低于 17。具体如何集成请参考 Higress 官方 BLOG [如何使用 Higress Admin SDK 进行配置管理](https://higress.io/zh-cn/blog/admin-sdk-intro)。
+
+## 2 Higress Controller
+
+Higress Controller（控制器） 是 Higress 的核心组件，其功能主要是实现 Higress 网关的服务发现、动态配置管理，以及动态下发配置给数据面。Higress Controller 内部包含两个子组件：Discovery 和  Higress Core。
+
+### 2.1 Discovery 组件
+
+Discovery 组件（Istio Pilot-Discovery）是 Istio 的核心组件，负责服务发现、配置管理、证书签发、控制面和数据面之间的通讯和配置下发等。Discovery 内部结构比较复杂，本文只介绍 Discovery 配置管理和服务发现的基本原理，其核心功能的详细介绍可以参考赵化冰老师的 BLOG [Istio Pilot 组件介绍](https://www.zhaohuabing.com/post/2019-10-21-pilot-discovery-code-analysis/)。
+Discovery 将 Kubernetes Service、Gateway API 配置等转换成 Istio 配置，然后将所有 Istio 配置合并转成符合 xDS 接口规范的数据结构，通过 GRPC 下发到数据面的 Envoy。其工作原理如下图：
+
+![img](./images/img_02_02.png)
+
+#### 2.1.1 Config Controller
+
+Discovery 为了更好管理 Istio 配置来源，提供 `Config Controller` 用于管理各种配置来源，目前支持 4 种类型的 `Config Controller`：
+
+- Kubernetes：使用 Kubernetes 作为配置信息来源，该方式的直接依赖 Kubernetes 强大的 CRD 机制来存储配置信息，简单方便，是 Istio 最开始使用的配置信息存储方案, 其中包括 `Kubernetes Controller` 和 `Gateway API Controller` 两个实现。
+- MCP（Mesh Configuration Protocol）：使用 Kubernetes 存储配置数据导致了 Istio 和 Kubernetes 的耦合，限制了 Istio 在非 Kubernetes 环境下的运用。为了解决该耦合，Istio 社区提出了 MCP。
+- Memory：一个基于内存的 Config Controller 实现，主要用于测试。
+- File：一个基于文件的 Config Controller 实现，主要用于测试。
+
+1. Istio 配置
+
+Istio 配置包括：`Gateway`、`VirtualService`、`DestinationRule`、`ServiceEntry`、`EnvoyFilter`、`WasmPlugin`、`WorkloadEntry`、`WorkloadGroup` 等，可以参考 Istio 官方文档[流量管理](https://istio.io/latest/zh/docs/reference/config/networking/)了解更多配置信息。
+
+2. Gateway API 配置
+
+Gateway API 配置包括：`GatewayClass`、`Gateway`、`HttpRoute`、`TCPRoute`、`GRPCRoute` 等, 可以参考 Gateway API 官方文档 [Gateway API](https://gateway-api.sigs.k8s.io/api-types/gateway/) 了解更多配置信息。
+
+3. MCP over xDS
+
+Discovery 作为 MCP Client，任何实现了 MCP 协议的 Server 都可以通过 MCP 协议向 Discovery 下发配置信息，从而消除了 Istio 和 Kubernetes 之间的耦合, 同时使 Istio 的配置信息处理更加灵活和可扩展。
+同时 MCP 是一种基于 xDS 协议的配置管理协议，Higress Core 通过实现 MCP 协议，使 Higress Core 成为 Discovery 的 Istio 配置来源。
+
+4. Config Controller 来源配置
+
+在 `higress-system` 命名空间中，名为 `higress-config` 的 Configmap 中，`mesh` 配置项包含一个 `configSources` 属性用于配置来源。其 Configmap 部分配置项如下：
+
+```yaml
+apiVersion: v1
+kind: ConfigMap
+metadata:
+  name: higress-config
+  namespace: higress-system
+data:
+  mesh: |-
+    accessLogEncoding: TEXT
+    ...
+    configSources:
+    - address: xds://127.0.0.1:15051
+    - address: k8s://
+    ...
+  meshNetworks: "networks: {}"
+```
+
+#### 2.1.2 Service Controller
+
+`Service Controller` 用于管理各种 `Service Registry`，提供服务发现数据，目前 Istio 支持的 `Service Registry` 主要包括：
+
+- Kubernetes：对接 Kubernetes Registry，可以将 Kubernetes 中定义的 Service 和 Endpoint 采集到 Istio 中。
+- Memory：一个基于内存的 Service Controller 实现，主要用于测试。
+
+### 2.2 Higress Core 组件
+
+Higress Core 核心逻辑如下图：
+
+![img](./images/img_02_03.png)
+
+
+Higress Core 内部包含两个核心子组件: Ingress Config 和 Cert Server。
+
+#### 2.2.1 Ingress Config
+
+Ingress Config 包含 6 个控制器，各自负责不同的功能：
+
+- Ingress Controller：监听 Ingress 资源，将 Ingress 转换为 Istio 的 Gateway、VirtualService、DestinationRule 等资源。
+- Gateway Controller：监听 Gateway、VirtualService、DestinationRule 等资源。
+- McpBridge Controller：根据 McpBridge 的配置，将来自 Nacos、Eureka、Consul、Zookeeper 等外部注册中心或 DNS 的服务信息转换成 Istio ServiceEntry 资源。
+- Http2Rpc Controller：监听 Http2Rpc 资源，实现 HTTP 协议到 RPC 协议的转换。用户可以通过配置协议转换，将 RPC 服务以 HTTP 接口的形式暴露，从而使用 HTTP 请求调用 RPC 接口。
+- WasmPlugin Controller：监听 WasmPlugin 资源，将 Higress WasmPlugin 转化为 Istio WasmPlugin。Higress WasmPlugin 在 Istio WasmPlugin 的基础上进行了扩展，支持全局、路由、域名、服务级别的配置。
+- ConfigmapMgr：监听 Higress 的全局配置 `higress-config` ConfigMap，可以根据 tracing、gzip 等配置构造 EnvoyFilter。
+
+#### 2.2.2 Cert Server
+
+Cert Server 管理 Secret 资源和证书自动签发。
+
+## 3 Higress Gateway
+
+Higress Gateway 内部包含两个子组件：Pilot Agent 和 Envoy。Pilot Agent 主要负责 Envoy 的启动和配置，同时代理 Envoy xDS 请求到 Discovery。 Envoy 作为数据面，负责接收控制面的配置下发，并代理请求到业务服务。 Pilot Agent 和 Envoy 之间通讯协议是使用 xDS 协议， 通过 Unix Domain Socket（UDS）进行通信。
+Envoy 核心架构如下图：
+
+![img](./images/img_02_04.png)
+
+### 1 Envoy 核心组件
+
+- 下游（Downstream）:
+  下游是 Envoy 的客户端，它们负责发起请求并接收 Envoy 的响应。下游通常是最终用户的设备或服务，它们通过 Envoy 代理与后端服务进行通信。
+
+- 上游（Upstream）:
+  上游是 Envoy 的后端服务器，它们接收 Envoy 代理的连接和请求。上游提供服务或数据，对来自下游客户端的请求进行处理并返回响应。
+
+- 监听器（Listener）:
+  监听器是可以接受来自下游客户端连接的网络地址（如 IP 地址和端口，Unix Domain Socket 等）。Envoy 支持在单个进程中配置任意数量的监听器。监听器可以通过 `Listener Discovery Service（LDS）`来动态发现和更新。
+
+- 路由（Router）:
+  路由器是 Envoy 中连接下游和上游的桥梁。它负责决定如何将监听器接收到的请求路由到适当的集群。路由器根据配置的路由规则，如路径、HTTP 标头 等，来确定请求的目标集群，从而实现精确的流量控制和路由。路由器可以通过 `Route Discovery Service（RDS）`来动态发现和更新。
+
+- 集群（Cluster）:
+  集群是一组逻辑上相似的服务提供者的集合。集群成员的选择由负载均衡策略决定，确保请求能够均匀或按需分配到不同的服务实例。集群可以通过 `Cluster Discovery Service（CDS）`来动态发现和更新。
+
+- 端点（Endpoint）:
+  端点是上游集群中的具体服务实例，可以是 IP 地址和端口号的组合。端点可以通过 `Endpoint Discovery Service（EDS）`来动态发现和更新。
+
+- SSL/TLS:
+  Envoy 可以通过 `Secret Discovery Service (SDS)` 动态获取监听器和集群所需的 TLS 证书、私钥以及信任的根证书和撤销机制等配置信息。
+
+通过这些组件的协同工作，Envoy 能够高效地处理网络请求，提供流量管理、负载均衡、服务发现和动态路由等关键功能。
+要详细了解 Envoy 的工作原理，可以参考[Envoy 官方文档](https://www.envoyproxy.io/docs/envoy/latest/intro/intro)，最佳的方式可以通过一个请求通过 [Envoy 代理的生命周期](https://www.envoyproxy.io/docs/envoy/latest/intro/life_of_a_request)事件的过程来理解 Envoy 的工作原理。
+
+
+## 参考
+
+- [1] [Istio Pilot 组件介绍](https://www.zhaohuabing.com/post/2019-10-21-pilot-discovery-code-analysis/)
+- [2] [Istio 服务注册插件机制代码解析](https://www.zhaohuabing.com/post/2019-02-18-pilot-service-registry-code-analysis/)
+- [3] [Istio Pilot代码深度解析](https://www.zhaohuabing.com/post/2019-10-21-pilot-discovery-code-analysis/)
+- [4] [Envoy 官方文档](https://www.envoyproxy.io/docs/envoy/latest/intro/intro)
--- a/docs/images/img_02_01.png
+++ b/docs/images/img_02_01.png
--- a/docs/images/img_02_02.png
+++ b/docs/images/img_02_02.png
--- a/docs/images/img_02_03.png
+++ b/docs/images/img_02_03.png
--- a/docs/images/img_02_04.png
+++ b/docs/images/img_02_04.png
--- a/envoy/envoy
+++ b/envoy/envoy
--- a/helm/core/Chart.yaml
+++ b/helm/core/Chart.yaml
@@ -1,5 +1,5 @@
 apiVersion: v2
-appVersion: 2.0.3
+appVersion: 2.0.6-rc.2
 description: Helm chart for deploying higress gateways
 icon: https://higress.io/img/higress_logo_small.png
 home: http://higress.io/
@@ -10,4 +10,4 @@ name: higress-core
 sources:
 - http://github.com/alibaba/higress
 type: application
-version: 2.0.3
+version: 2.0.6-rc.2
--- a/helm/core/templates/_pod.tpl
+++ b/helm/core/templates/_pod.tpl
@@ -7,9 +7,6 @@ Rendering the pod template of gateway component.
 template:
  metadata:
    annotations:
-    {{- if .Values.global.enableHigressIstio }}
-      "enableHigressIstio": "true"
-    {{- end }}
    {{- if .Values.gateway.podAnnotations }}
      {{- toYaml .Values.gateway.podAnnotations | nindent 6 }}
    {{- end }}
@@ -268,11 +265,7 @@ template:
    {{- end }}
    - name: higress-ca-root-cert
      configMap:
-    {{- if .Values.global.enableHigressIstio }}
-        name: istio-ca-root-cert
-    {{- else }}
        name: higress-ca-root-cert
-    {{- end }}
    - name: config
      configMap:
        name: higress-config
--- a/helm/core/templates/configmap.yaml
+++ b/helm/core/templates/configmap.yaml
@@ -9,7 +9,7 @@
    accessLogFile: "/dev/stdout"
    {{- end }}
    ingressControllerMode: "OFF"
-    accessLogFormat: '{"authority":"%REQ(X-ENVOY-ORIGINAL-HOST?:AUTHORITY)%","bytes_received":"%BYTES_RECEIVED%","bytes_sent":"%BYTES_SENT%","downstream_local_address":"%DOWNSTREAM_LOCAL_ADDRESS%","downstream_remote_address":"%DOWNSTREAM_REMOTE_ADDRESS%","duration":"%DURATION%","istio_policy_status":"%DYNAMIC_METADATA(istio.mixer:status)%","method":"%REQ(:METHOD)%","path":"%REQ(X-ENVOY-ORIGINAL-PATH?:PATH)%","protocol":"%PROTOCOL%","request_id":"%REQ(X-REQUEST-ID)%","requested_server_name":"%REQUESTED_SERVER_NAME%","response_code":"%RESPONSE_CODE%","response_flags":"%RESPONSE_FLAGS%","route_name":"%ROUTE_NAME%","start_time":"%START_TIME%","trace_id":"%REQ(X-B3-TRACEID)%","upstream_cluster":"%UPSTREAM_CLUSTER%","upstream_host":"%UPSTREAM_HOST%","upstream_local_address":"%UPSTREAM_LOCAL_ADDRESS%","upstream_service_time":"%RESP(X-ENVOY-UPSTREAM-SERVICE-TIME)%","upstream_transport_failure_reason":"%UPSTREAM_TRANSPORT_FAILURE_REASON%","user_agent":"%REQ(USER-AGENT)%","x_forwarded_for":"%REQ(X-FORWARDED-FOR)%","response_code_details":"%RESPONSE_CODE_DETAILS%"}
+    accessLogFormat: '{"ai_log":"%FILTER_STATE(wasm.ai_log:PLAIN)%","authority":"%REQ(X-ENVOY-ORIGINAL-HOST?:AUTHORITY)%","bytes_received":"%BYTES_RECEIVED%","bytes_sent":"%BYTES_SENT%","downstream_local_address":"%DOWNSTREAM_LOCAL_ADDRESS%","downstream_remote_address":"%DOWNSTREAM_REMOTE_ADDRESS%","duration":"%DURATION%","istio_policy_status":"%DYNAMIC_METADATA(istio.mixer:status)%","method":"%REQ(:METHOD)%","path":"%REQ(X-ENVOY-ORIGINAL-PATH?:PATH)%","protocol":"%PROTOCOL%","request_id":"%REQ(X-REQUEST-ID)%","requested_server_name":"%REQUESTED_SERVER_NAME%","response_code":"%RESPONSE_CODE%","response_flags":"%RESPONSE_FLAGS%","route_name":"%ROUTE_NAME%","start_time":"%START_TIME%","trace_id":"%REQ(X-B3-TRACEID)%","upstream_cluster":"%UPSTREAM_CLUSTER%","upstream_host":"%UPSTREAM_HOST%","upstream_local_address":"%UPSTREAM_LOCAL_ADDRESS%","upstream_service_time":"%RESP(X-ENVOY-UPSTREAM-SERVICE-TIME)%","upstream_transport_failure_reason":"%UPSTREAM_TRANSPORT_FAILURE_REASON%","user_agent":"%REQ(USER-AGENT)%","x_forwarded_for":"%REQ(X-FORWARDED-FOR)%","response_code_details":"%RESPONSE_CODE_DETAILS%"}

    '
    dnsRefreshRate: 200s
@@ -20,11 +20,7 @@
    # When processing a leaf namespace Istio will search for declarations in that namespace first
    # and if none are found it will search in the root namespace. Any matching declaration found in the root namespace
    # is processed as if it were declared in the leaf namespace.
-    {{- if .Values.global.enableHigressIstio }}
-    rootNamespace: {{ .Values.meshConfig.rootNamespace | default .Values.global.istioNamespace }}
-    {{- else }}
    rootNamespace: {{ .Release.Namespace }}
-    {{- end }}

    configSources:
      - address: "xds://127.0.0.1:15051"
@@ -85,12 +81,8 @@
      discoveryAddress: {{ printf "istiod.%s.svc" .Release.Namespace }}:15012
      {{- end }}
      {{- else }}
-      {{- if .Values.global.enableHigressIstio }}
-      discoveryAddress: {{ printf "istiod.%s.svc" .Values.global.istioNamespace }}:15012
-      {{- else }}
      discoveryAddress: {{ include "controller.name" . }}.{{.Release.Namespace}}.svc:15012
      {{- end }}
-      {{- end }}
      proxyStatsMatcher:
        inclusionRegexps:
        - ".*"
--- a/helm/core/templates/controller-deployment.yaml
+++ b/helm/core/templates/controller-deployment.yaml
@@ -96,7 +96,6 @@ spec:
          volumeMounts:
          - name: log
            mountPath: /var/log
-{{- if not .Values.global.enableHigressIstio }}
        - name: discovery
          image: "{{ .Values.pilot.hub | default .Values.global.hub }}/{{ .Values.pilot.image | default "pilot" }}:{{ .Values.pilot.tag | default .Chart.AppVersion }}"
 {{- if .Values.global.imagePullPolicy }}
@@ -137,6 +136,10 @@ spec:
            periodSeconds: 3
            timeoutSeconds: 5
          env:
+          - name: ENABLE_PUSH_ALL_MCP_CLUSTERS
+            value: "{{ .Values.global.enablePushAllMCPClusters }}"
+          - name: PILOT_ENABLE_LDS_CACHE
+            value: "{{ .Values.global.enableLDSCache }}"
          - name: PILOT_ENABLE_QUIC_LISTENERS
            value: "true"
          - name: VALIDATION_WEBHOOK_CONFIG_NAME
@@ -229,10 +232,8 @@ spec:
            value: "false"
          - name: PILOT_ENABLE_GATEWAY_API_DEPLOYMENT_CONTROLLER
            value: "false"
-          {{- if not .Values.global.enableHigressIstio }}
          - name: CUSTOM_CA_CERT_NAME
            value: "higress-ca-root-cert"
-          {{- end }}
          {{- if not (or .Values.global.local .Values.global.kind) }}
          resources:
 {{- if .Values.pilot.resources }}
@@ -269,7 +270,6 @@ spec:
          - name: extracacerts
            mountPath: /cacerts
          {{- end }}
-{{- end }}
      {{- with .Values.controller.nodeSelector }}
      nodeSelector:
        {{- toYaml . | nindent 8 }}
@@ -285,7 +285,6 @@ spec:
      volumes:
      - name: log
        emptyDir: {}
-      {{- if not .Values.global.enableHigressIstio }}
      - name: config
        configMap:
          name: higress-config
@@ -317,4 +316,3 @@ spec:
        configMap:
          name: pilot-jwks-extra-cacerts{{- if not (eq .Values.revision "") }}-{{ .Values.revision }}{{- end }}
  {{- end }}
-      {{- end }}
--- a/helm/core/templates/controller-service.yaml
+++ b/helm/core/templates/controller-service.yaml
@@ -9,7 +9,6 @@ spec:
  type: {{ .Values.controller.service.type }}
  ports:
    {{- toYaml .Values.controller.ports | nindent 4 }}
-    {{- if not .Values.global.enableHigressIstio }}
    - port: 15010
      name: grpc-xds # plaintext
      protocol: TCP
@@ -23,6 +22,5 @@ spec:
    - port: 15014
      name: http-monitoring # prometheus stats
      protocol: TCP
-    {{- end }}
  selector:
    {{- include "controller.selectorLabels" . | nindent 4 }}
--- a/helm/core/templates/daemonset.yaml
+++ b/helm/core/templates/daemonset.yaml
@@ -1,7 +1,8 @@
 {{- if eq .Values.gateway.kind "DaemonSet" -}}
 {{- $o11y := .Values.global.o11y  }}
-{{- $unprivilegedPortSupported := true }}
-{{- range $index, $node := (lookup "v1" "Node" "default" "").items }}
+{{- if eq .Values.gateway.unprivilegedPortSupported nil -}}
+  {{- $unprivilegedPortSupported := true }}
+  {{- range $index, $node := (lookup "v1" "Node" "default" "").items }}
    {{- $kernelVersion := $node.status.nodeInfo.kernelVersion }}
    {{- if $kernelVersion }}
      {{- $kernelVersion = regexFind "^(\\d+\\.\\d+\\.\\d+)" $kernelVersion }}
@@ -9,8 +10,9 @@
      {{- $unprivilegedPortSupported = false }}
      {{- end }}
    {{- end }}
+  {{- end -}}
+  {{- $_ := set .Values.gateway "unprivilegedPortSupported" $unprivilegedPortSupported -}}
 {{- end -}}
-{{- $_ := set .Values.gateway "unprivilegedPortSupported" $unprivilegedPortSupported -}}

 apiVersion: apps/v1
 kind: DaemonSet
--- a/helm/core/templates/deployment.yaml
+++ b/helm/core/templates/deployment.yaml
@@ -1,6 +1,7 @@
 {{- if eq .Values.gateway.kind "Deployment" -}}
-{{- $unprivilegedPortSupported := true }}
-{{- range $index, $node := (lookup "v1" "Node" "default" "").items }}
+{{- if eq .Values.gateway.unprivilegedPortSupported nil -}}
+  {{- $unprivilegedPortSupported := true }}
+  {{- range $index, $node := (lookup "v1" "Node" "default" "").items }}
    {{- $kernelVersion := $node.status.nodeInfo.kernelVersion }}
    {{- if $kernelVersion }}
      {{- $kernelVersion = regexFind "^(\\d+\\.\\d+\\.\\d+)" $kernelVersion }}
@@ -8,8 +9,9 @@
      {{- $unprivilegedPortSupported = false }}
      {{- end }}
    {{- end }}
+  {{- end -}}
+  {{- $_ := set .Values.gateway "unprivilegedPortSupported" $unprivilegedPortSupported -}}
 {{- end -}}
-{{- $_ := set .Values.gateway "unprivilegedPortSupported" $unprivilegedPortSupported -}}

 apiVersion: apps/v1
 kind: Deployment
--- a/helm/core/values.yaml
+++ b/helm/core/values.yaml
@@ -3,14 +3,16 @@ global:
  enableH3: false
  enableIPv6: false
  enableProxyProtocol: false
-  liteMetrics: true
+  enableLDSCache: true
+  enablePushAllMCPClusters: true
+  liteMetrics: false
  xdsMaxRecvMsgSize: "104857600"
  defaultUpstreamConcurrencyThreshold: 10000
  enableSRDS: true
  onDemandRDS: false
  hostRDSMergeSubset: false
  onlyPushRouteCluster: true
-  # IngressClass filters which ingress resources the higress controller watches.
+  # -- IngressClass filters which ingress resources the higress controller watches.
  # The default ingress class is higress.
  # There are some special cases for special ingress class.
  # 1. When the ingress class is set as nginx, the higress controller will watch ingress
@@ -18,28 +20,38 @@ global:
  # 2. When the ingress class is set empty, the higress controller will watch all ingress
  # resources in the k8s cluster.
  ingressClass: "higress"
+  # -- If not empty, Higress Controller will only watch resources in the specified namespace.
+  # When isolating different business systems using K8s namespace,
+  # if each namespace requires a standalone gateway instance,
+  # this parameter can be used to confine the Ingress watching of Higress within the given namespace.
  watchNamespace: ""
+  # -- Whether to disable HTTP/2 in ALPN
  disableAlpnH2: false
+  # -- If true, Higress Controller will update the status field of Ingress resources.
+  # When migrating from Nginx Ingress, in order to avoid status field of Ingress objects being overwritten,
+  # this parameter needs to be set to false,
+  # so Higress won't write the entry IP to the status field of the corresponding Ingress object.
  enableStatus: true
-  # whether to use autoscaling/v2 template for HPA settings
+  # -- whether to use autoscaling/v2 template for HPA settings
  # for internal usage only, not to be configured by users.
  autoscalingv2API: true
-  local: false # When deploying to a local cluster (e.g.: kind cluster), set this to true.
+  # -- When deploying to a local cluster (e.g.: kind cluster), set this to true.
+  local: false
  kind: false # Deprecated. Please use "global.local" instead. Will be removed later.
+  # -- If true, Higress Controller will monitor istio resources as well
  enableIstioAPI: true
+  # -- If true, Higress Controller will monitor Gateway API resources as well
  enableGatewayAPI: false
-  # Deprecated
-  enableHigressIstio: false
-  # Used to locate istiod.
+  # -- Used to locate istiod.
  istioNamespace: istio-system
-  # enable pod disruption budget for the control plane, which is used to
+  # -- enable pod disruption budget for the control plane, which is used to
  # ensure Istio control plane components are gradually upgraded or recovered.
  defaultPodDisruptionBudget:
    enabled: false
    # The values aren't mutable due to a current PodDisruptionBudget limitation
    # minAvailable: 1

-  # A minimal set of requested resources to applied to all deployments so that
+  # -- A minimal set of requested resources to applied to all deployments so that
  # Horizontal Pod Autoscaler will be able to function (if set).
  # Each component can overwrite these default values by adding its own resources
  # block in the relevant section below and setting the desired resources values.
@@ -51,16 +63,16 @@ global:
    #   cpu: 100m
    #   memory: 128Mi

-  # Default hub for Istio images.
+  # -- Default hub for Istio images.
  # Releases are published to docker hub under 'istio' project.
  # Dev builds from prow are on gcr.io
  hub: higress-registry.cn-hangzhou.cr.aliyuncs.com/higress

-  # Specify image pull policy if default behavior isn't desired.
+  # -- Specify image pull policy if default behavior isn't desired.
  # Default behavior: latest images will be Always else IfNotPresent.
  imagePullPolicy: ""

-  # ImagePullSecrets for all ServiceAccount, list of secrets in the same namespace
+  # -- ImagePullSecrets for all ServiceAccount, list of secrets in the same namespace
  # to use for pulling any images in pods that reference this ServiceAccount.
  # For components that don't use ServiceAccounts (i.e. grafana, servicegraph, tracing)
  # ImagePullSecrets will be added to the corresponding Deployment(StatefulSet) objects.
@@ -68,14 +80,14 @@ global:
  imagePullSecrets: []
  # - private-registry-key

-  # Enabled by default in master for maximising testing.
+  # -- Enabled by default in master for maximising testing.
  istiod:
    enableAnalysis: false

-  # To output all istio components logs in json format by adding --log_as_json argument to each container argument
+  # -- To output all istio components logs in json format by adding --log_as_json argument to each container argument
  logAsJson: false

-  # Comma-separated minimum per-scope logging level of messages to output, in the form of <scope>:<level>,<scope>:<level>
+  # -- Comma-separated minimum per-scope logging level of messages to output, in the form of <scope>:<level>,<scope>:<level>
  # The control plane has different scopes depending on component, but can configure default log level across all components
  # If empty, default scope and level will be used as configured in code
  logging:
@@ -83,11 +95,11 @@ global:

  omitSidecarInjectorConfigMap: false

-  # Whether to restrict the applications namespace the controller manages;
+  # -- Whether to restrict the applications namespace the controller manages;
  # If not set, controller watches all namespaces
  oneNamespace: false

-  # Configure whether Operator manages webhook configurations. The current behavior
+  # -- Configure whether Operator manages webhook configurations. The current behavior
  # of Istiod is to manage its own webhook configurations.
  # When this option is set as true, Istio Operator, instead of webhooks, manages the
  # webhook configurations. When this option is set as false, webhooks manage their
@@ -106,7 +118,7 @@ global:
  #- global
  #- "{{ valueOrDefault .DeploymentMeta.Namespace \"default\" }}.global"

-  # Kubernetes >=v1.11.0 will create two PriorityClass, including system-cluster-critical and
+  # -- Kubernetes >=v1.11.0 will create two PriorityClass, including system-cluster-critical and
  # system-node-critical, it is better to configure this in order to make sure your Istio pods
  # will not be killed because of low priority class.
  # Refer to https://kubernetes.io/docs/concepts/configuration/pod-priority-preemption/#priorityclass
@@ -116,18 +128,18 @@ global:
  proxy:
    image: proxyv2

-    # This controls the 'policy' in the sidecar injector.
+    # -- This controls the 'policy' in the sidecar injector.
    autoInject: enabled

-    # CAUTION: It is important to ensure that all Istio helm charts specify the same clusterDomain value
+    # -- CAUTION: It is important to ensure that all Istio helm charts specify the same clusterDomain value
    # cluster domain. Default value is "cluster.local".
    clusterDomain: "cluster.local"

-    # Per Component log level for proxy, applies to gateways and sidecars. If a component level is
+    # -- Per Component log level for proxy, applies to gateways and sidecars. If a component level is
    # not set, then the global "logLevel" will be used.
    componentLogLevel: "misc:error"

-    # If set, newly injected sidecars will have core dumps enabled.
+    # -- If set, newly injected sidecars will have core dumps enabled.
    enableCoreDump: false

    # istio ingress capture allowlist
@@ -136,7 +148,7 @@ global:
    excludeInboundPorts: ""
    includeInboundPorts: "*"

-    # istio egress capture allowlist
+    # -- istio egress capture allowlist
    # https://istio.io/docs/tasks/traffic-management/egress.html#calling-external-services-directly
    # example: includeIPRanges: "172.30.0.0/16,172.20.0.0/16"
    # would only capture egress traffic on those two IP Ranges, all other outbound traffic would
@@ -146,29 +158,29 @@ global:
    includeOutboundPorts: ""
    excludeOutboundPorts: ""

-    # Log level for proxy, applies to gateways and sidecars.
+    # -- Log level for proxy, applies to gateways and sidecars.
    # Expected values are: trace|debug|info|warning|error|critical|off
    logLevel: warning

-    #If set to true, istio-proxy container will have privileged securityContext
+    # -- If set to true, istio-proxy container will have privileged securityContext
    privileged: false

-    # The number of successive failed probes before indicating readiness failure.
+    # -- The number of successive failed probes before indicating readiness failure.
    readinessFailureThreshold: 30

-    # The number of successive successed probes before indicating readiness success.
+    # -- The number of successive successed probes before indicating readiness success.
    readinessSuccessThreshold: 30

-    # The initial delay for readiness probes in seconds.
+    # -- The initial delay for readiness probes in seconds.
    readinessInitialDelaySeconds: 1

-    # The period between readiness probes.
+    # -- The period between readiness probes.
    readinessPeriodSeconds: 2

-    # The readiness timeout seconds
+    # -- The readiness timeout seconds
    readinessTimeoutSeconds: 3

-    # Resources for the sidecar.
+    # -- Resources for the sidecar.
    resources:
      requests:
        cpu: 100m
@@ -177,18 +189,18 @@ global:
        cpu: 2000m
        memory: 1024Mi

-    # Default port for Pilot agent health checks. A value of 0 will disable health checking.
+    # -- Default port for Pilot agent health checks. A value of 0 will disable health checking.
    statusPort: 15020

-    # Specify which tracer to use. One of: lightstep, datadog, stackdriver.
+    # -- Specify which tracer to use. One of: lightstep, datadog, stackdriver.
    # If using stackdriver tracer outside GCP, set env GOOGLE_APPLICATION_CREDENTIALS to the GCP credential file.
    tracer: ""

-    # Controls if sidecar is injected at the front of the container list and blocks the start of the other containers until the proxy is ready
+    # -- Controls if sidecar is injected at the front of the container list and blocks the start of the other containers until the proxy is ready
    holdApplicationUntilProxyStarts: false

  proxy_init:
-    # Base name for the proxy_init container, used to configure iptables.
+    # -- Base name for the proxy_init container, used to configure iptables.
    image: proxyv2
    resources:
      limits:
@@ -198,7 +210,7 @@ global:
        cpu: 10m
        memory: 10Mi

-  # configure remote pilot and istiod service and endpoint
+  # -- configure remote pilot and istiod service and endpoint
  remotePilotAddress: ""

  ##############################################################################################
@@ -206,20 +218,20 @@ global:
  # make sure they are consistent across your Istio helm charts                                #
  ##############################################################################################

-  # The customized CA address to retrieve certificates for the pods in the cluster.
+  # -- The customized CA address to retrieve certificates for the pods in the cluster.
  # CSR clients such as the Istio Agent and ingress gateways can use this to specify the CA endpoint.
  # If not set explicitly, default to the Istio discovery address.
  caAddress: ""

-  # Configure a remote cluster data plane controlled by an external istiod.
+  # -- Configure a remote cluster data plane controlled by an external istiod.
  # When set to true, istiod is not deployed locally and only a subset of the other
  # discovery charts are enabled.
  externalIstiod: false

-  #  Configure a remote cluster as the config cluster for an external istiod.
+  # -- Configure a remote cluster as the config cluster for an external istiod.
  configCluster: false

-  # Configure the policy for validating JWT.
+  # -- Configure the policy for validating JWT.
  # Currently, two options are supported: "third-party-jwt" and "first-party-jwt".
  jwtPolicy: "third-party-jwt"

@@ -241,7 +253,7 @@ global:
  # of migration TBD, and it may be a disruptive operation to change the Mesh
  # ID post-install.
  #
-  # If the mesh admin does not specify a value, Istio will use the value of the
+  # -- If the mesh admin does not specify a value, Istio will use the value of the
  # mesh's Trust Domain. The best practice is to select a proper Trust Domain
  # value.
  meshID: ""
@@ -275,68 +287,69 @@ global:
  #
  meshNetworks: {}

-  # Use the user-specified, secret volume mounted key and certs for Pilot and workloads.
+  # -- Use the user-specified, secret volume mounted key and certs for Pilot and workloads.
  mountMtlsCerts: false

  multiCluster:
-    # Set to true to connect two kubernetes clusters via their respective
+    # -- Set to true to connect two kubernetes clusters via their respective
    # ingressgateway services when pods in each cluster cannot directly
    # talk to one another. All clusters should be using Istio mTLS and must
    # have a shared root CA for this model to work.
    enabled: true
-    # Should be set to the name of the cluster this installation will run in. This is required for sidecar injection
+    # -- Should be set to the name of the cluster this installation will run in. This is required for sidecar injection
    # to properly label proxies
    clusterName: ""

-  # Network defines the network this cluster belong to. This name
+  # -- Network defines the network this cluster belong to. This name
  # corresponds to the networks in the map of mesh networks.
  network: ""

-  # Configure the certificate provider for control plane communication.
+  # -- Configure the certificate provider for control plane communication.
  # Currently, two providers are supported: "kubernetes" and "istiod".
  # As some platforms may not have kubernetes signing APIs,
  # Istiod is the default
  pilotCertProvider: istiod

  sds:
-    # The JWT token for SDS and the aud field of such JWT. See RFC 7519, section 4.1.3.
+    # -- The JWT token for SDS and the aud field of such JWT. See RFC 7519, section 4.1.3.
    # When a CSR is sent from Istio Agent to the CA (e.g. Istiod), this aud is to make sure the
    # JWT is intended for the CA.
    token:
      aud: istio-ca

  sts:
-    # The service port used by Security Token Service (STS) server to handle token exchange requests.
+    # -- The service port used by Security Token Service (STS) server to handle token exchange requests.
    # Setting this port to a non-zero value enables STS server.
    servicePort: 0

-  # Configuration for each of the supported tracers
+  # -- Configuration for each of the supported tracers
  tracer:
-    # Configuration for envoy to send trace data to LightStep.
+    # -- Configuration for envoy to send trace data to LightStep.
    # Disabled by default.
    # address: the <host>:<port> of the satellite pool
    # accessToken: required for sending data to the pool
    #
    datadog:
-      # Host:Port for submitting traces to the Datadog agent.
+      # -- Host:Port for submitting traces to the Datadog agent.
      address: "$(HOST_IP):8126"
    lightstep:
-      address: "" # example: lightstep-satellite:443
-      accessToken: "" # example: abcdefg1234567
+      # -- example: lightstep-satellite:443
+      address: ""
+      # -- example: abcdefg1234567
+      accessToken: ""
    stackdriver:
-      # enables trace output to stdout.
+      # -- enables trace output to stdout.
      debug: false
-      # The global default max number of message events per span.
+      # -- The global default max number of message events per span.
      maxNumberOfMessageEvents: 200
-      # The global default max number of annotation events per span.
+      # -- The global default max number of annotation events per span.
      maxNumberOfAnnotations: 200
-      # The global default max number of attributes per span.
+      # -- The global default max number of attributes per span.
      maxNumberOfAttributes: 200
-  # Use the Mesh Control Protocol (MCP) for configuring Istiod. Requires an MCP source.
-
+  # -- Use the Mesh Control Protocol (MCP) for configuring Istiod. Requires an MCP source.
  useMCP: false

-  # Observability (o11y) configurations
+  # -- Observability (o11y) configurations
  o11y:
    enabled: false
    promtail:
@@ -350,7 +363,7 @@ global:
          memory: 2Gi
      securityContext: {}

-  # The name of the CA for workload certificates.
+  # -- The name of the CA for workload certificates.
  # For example, when caName=GkeWorkloadCertificate, GKE workload certificates
  # will be used as the certificates for workloads.
  # The default value is "" and when caName="", the CA will be configured by other
@@ -359,7 +372,7 @@ global:
 hub: higress-registry.cn-hangzhou.cr.aliyuncs.com/higress

 clusterName: ""
-# meshConfig defines runtime configuration of components, including Istiod and istio-agent behavior
+# -- meshConfig defines runtime configuration of components, including Istiod and istio-agent behavior
 # See https://istio.io/docs/reference/config/istio.mesh.v1alpha1/ for all available options
 meshConfig:
  enablePrometheusMerge: true
@@ -370,14 +383,13 @@ meshConfig:
  # and gradual adoption by setting capture only on specific workloads. It also allows
  # VMs to use other DNS options, like dnsmasq or unbound.

-  # The namespace to treat as the administrative root namespace for Istio configuration.
+  # -- The namespace to treat as the administrative root namespace for Istio configuration.
  # When processing a leaf namespace Istio will search for declarations in that namespace first
  # and if none are found it will search in the root namespace. Any matching declaration found in the root namespace
  # is processed as if it were declared in the leaf namespace.
-
  rootNamespace:

-  # The trust domain corresponds to the trust root of a system
+  # -- The trust domain corresponds to the trust root of a system
  # Refer to https://github.com/spiffe/spiffe/blob/master/standards/SPIFFE-ID.md#21-trust-domain
  trustDomain: "cluster.local"

@@ -391,56 +403,57 @@ meshConfig:

 gateway:
  name: "higress-gateway"
+  # -- Number of Higress Gateway pods
  replicas: 2
  image: gateway

  # -- Use a `DaemonSet` or `Deployment`
  kind: Deployment

-  # The number of successive failed probes before indicating readiness failure.
+  # -- The number of successive failed probes before indicating readiness failure.
  readinessFailureThreshold: 30

-  # The number of successive successed probes before indicating readiness success.
+  # -- The number of successive successed probes before indicating readiness success.
  readinessSuccessThreshold: 1

-  # The initial delay for readiness probes in seconds.
+  # -- The initial delay for readiness probes in seconds.
  readinessInitialDelaySeconds: 1

-  # The period between readiness probes.
+  # -- The period between readiness probes.
  readinessPeriodSeconds: 2

-  # The readiness timeout seconds
+  # -- The readiness timeout seconds
  readinessTimeoutSeconds: 3

  hub: higress-registry.cn-hangzhou.cr.aliyuncs.com/higress
  tag: ""
-  # revision declares which revision this gateway is a part of
+  # -- revision declares which revision this gateway is a part of
  revision: ""

  rbac:
-    # If enabled, roles will be created to enable accessing certificates from Gateways. This is not needed
+    # -- If enabled, roles will be created to enable accessing certificates from Gateways. This is not needed
    # when using http://gateway-api.org/.
    enabled: true

  serviceAccount:
-    # If set, a service account will be created. Otherwise, the default is used
+    # -- If set, a service account will be created. Otherwise, the default is used
    create: true
-    # Annotations to add to the service account
+    # -- Annotations to add to the service account
    annotations: {}
-    # The name of the service account to use.
+    # -- The name of the service account to use.
    # If not set, the release name is used
    name: ""

-  # Pod environment variables
+  # -- Pod environment variables
  env: {}
  httpPort: 80
  httpsPort: 443
  hostNetwork: false

-  # Labels to apply to all resources
+  # -- Labels to apply to all resources
  labels: {}

-  # Annotations to apply to all resources
+  # -- Annotations to apply to all resources
  annotations: {}

  podAnnotations:
@@ -449,14 +462,15 @@ gateway:
    prometheus.io/path: "/stats/prometheus"
    sidecar.istio.io/inject: "false"

-  # Define the security context for the pod.
+  # -- Define the security context for the pod.
  # If unset, this will be automatically set to the minimum privileges required to bind to port 80 and 443.
  # On Kubernetes 1.22+, this only requires the `net.ipv4.ip_unprivileged_port_start` sysctl.
  securityContext: ~
  containerSecurityContext: ~
+  unprivilegedPortSupported: ~

  service:
-    # Type of service. Set to "None" to disable the service entirely
+    # -- Type of service. Set to "None" to disable the service entirely
    type: LoadBalancer
    ports:
      - name: http2
@@ -496,28 +510,29 @@ gateway:

  affinity: {}

-  # If specified, the gateway will act as a network gateway for the given network.
+  # -- If specified, the gateway will act as a network gateway for the given network.
  networkGateway: ""

  metrics:
-    # If true, create PodMonitor or VMPodScrape for gateway
+    # -- If true, create PodMonitor or VMPodScrape for gateway
    enabled: false
-    # provider group name for CustomResourceDefinition, can be monitoring.coreos.com or operator.victoriametrics.com
+    # -- provider group name for CustomResourceDefinition, can be monitoring.coreos.com or operator.victoriametrics.com
    provider: monitoring.coreos.com
    interval: ""
    scrapeTimeout: ""
    honorLabels: false
-    # for monitoring.coreos.com/v1.PodMonitor
+    # -- for monitoring.coreos.com/v1.PodMonitor
    metricRelabelings: []
    relabelings: []
-    # for operator.victoriametrics.com/v1beta1.VMPodScrape
+    # -- for operator.victoriametrics.com/v1beta1.VMPodScrape
    metricRelabelConfigs: []
    relabelConfigs: []
-    # some more raw podMetricsEndpoints spec
+    # -- some more raw podMetricsEndpoints spec
    rawSpec: {}

 controller:
  name: "higress-controller"
+  # -- Number of Higress Controller pods
  replicas: 1
  image: higress

@@ -541,12 +556,12 @@ controller:
    create: true

  serviceAccount:
-    # Specifies whether a service account should be created
+    # -- Specifies whether a service account should be created
    create: true
-    # Annotations to add to the service account
+    # -- Annotations to add to the service account
    annotations: {}
-    # The name of the service account to use.
-    # If not set and create is true, a name is generated using the fullname template
+    # -- The name of the service account to use.
+    # -- If not set and create is true, a name is generated using the fullname template
    name: ""

  podAnnotations: {}
@@ -602,7 +617,7 @@ controller:
    enabled: true
    email: ""

-## Discovery Settings
+## -- Discovery Settings
 pilot:
  autoscaleEnabled: false
  autoscaleMin: 1
@@ -614,11 +629,11 @@ pilot:
  hub: higress-registry.cn-hangzhou.cr.aliyuncs.com/higress
  tag: ""

-  # Can be a full hub/image:tag
+  # -- Can be a full hub/image:tag
  image: pilot
  traceSampling: 1.0

-  # Resources for a small pilot install
+  # -- Resources for a small pilot install
  resources:
    requests:
      cpu: 500m
@@ -633,21 +648,21 @@ pilot:
  cpu:
    targetAverageUtilization: 80

-  # if protocol sniffing is enabled for outbound
+  # -- if protocol sniffing is enabled for outbound
  enableProtocolSniffingForOutbound: true
-  # if protocol sniffing is enabled for inbound
+  # -- if protocol sniffing is enabled for inbound
  enableProtocolSniffingForInbound: true

  nodeSelector: {}
  podAnnotations: {}
  serviceAnnotations: {}

-  # You can use jwksResolverExtraRootCA to provide a root certificate
+  # -- You can use jwksResolverExtraRootCA to provide a root certificate
  # in PEM format. This will then be trusted by pilot when resolving
  # JWKS URIs.
  jwksResolverExtraRootCA: ""

-  # This is used to set the source of configuration for
+  # -- This is used to set the source of configuration for
  # the associated address in configSource, if nothing is specified
  # the default MCP is assumed.
  configSource:
@@ -655,21 +670,21 @@ pilot:

  plugins: []

-  # The following is used to limit how long a sidecar can be connected
+  # -- The following is used to limit how long a sidecar can be connected
  # to a pilot. It balances out load across pilot instances at the cost of
  # increasing system churn.
  keepaliveMaxServerConnectionAge: 30m

-  # Additional labels to apply to the deployment.
+  # -- Additional labels to apply to the deployment.
  deploymentLabels: {}

  ## Mesh config settings

-  # Install the mesh config map, generated from values.yaml.
+  # -- Install the mesh config map, generated from values.yaml.
  # If false, pilot wil use default values (by default) or user-supplied values.
  configMap: true

-  # Additional labels to apply on the pod level for monitoring and logging configuration.
+  # -- Additional labels to apply on the pod level for monitoring and logging configuration.
  podLabels: {}

 # Tracing config settings
@@ -685,7 +700,7 @@ tracing:
  # service: ""
  # port: 9411

-# Downstream config settings
+# -- Downstream config settings
 downstream:
  idleTimeout: 180
  maxRequestHeadersKb: 60
@@ -696,7 +711,7 @@ downstream:
    initialConnectionWindowSize: 1048576
  routeTimeout: 0

-# Upstream config settings
+# -- Upstream config settings
 upstream:
  idleTimeout: 10
  connectionBufferLimits: 10485760
--- a/helm/higress/Chart.lock
+++ b/helm/higress/Chart.lock
@@ -1,9 +1,9 @@
 dependencies:
 - name: higress-core
  repository: file://../core
-  version: 2.0.3
+  version: 2.0.6-rc.2
 - name: higress-console
  repository: https://higress.io/helm-charts/
-  version: 1.4.5
-digest: sha256:74b772113264168483961f5d0424459fd7359adc509a4b50400229581d7cddbf
-generated: "2024-11-08T14:06:51.871719+08:00"
+  version: 2.0.1
+digest: sha256:084449006a5b90bdffad7ef47fdfd02924412e67297bcce6d216efdc12c02acf
+generated: "2025-01-14T19:37:32.036755+08:00"
--- a/helm/higress/Chart.yaml
+++ b/helm/higress/Chart.yaml
@@ -1,5 +1,5 @@
 apiVersion: v2
-appVersion: 2.0.3
+appVersion: 2.0.6-rc.2
 description: Helm chart for deploying Higress gateways
 icon: https://higress.io/img/higress_logo_small.png
 home: http://higress.io/
@@ -12,9 +12,9 @@ sources:
 dependencies:
 - name: higress-core
  repository: "file://../core"
-  version: 2.0.3
+  version: 2.0.6-rc.2
 - name: higress-console
  repository: "https://higress.io/helm-charts/"
-  version: 1.4.5
+  version: 2.0.1
 type: application
-version: 2.0.3
+version: 2.0.6-rc.2
--- a/helm/higress/README.md
+++ b/helm/higress/README.md
@@ -1,57 +1,278 @@
-# Higress Helm Chart
-
-Installs the cloud-native gateway [Higress](http://higress.io/)
-
-## Get Repo Info
-
-```console
-helm repo add higress.io https://higress.io/helm-charts
-helm repo update
-```
-
-_See [helm repo](https://helm.sh/docs/helm/helm_repo/) for command documentation._
-
-## Installing the Chart
-
-To install the chart with the release name `higress`:
-
-```console
-helm install higress -n higress-system higress.io/higress --create-namespace --render-subchart-notes
-```
-
-## Uninstalling the Chart
-
-To uninstall/delete the higress deployment:
-
-```console
-helm delete higress -n higress-system
-```
-
-The command removes all the Kubernetes components associated with the chart and deletes the release.
-
-## Configuration
-
-| **Parameter** | **Description** | **Default** |
-|---|---|---|
-| **Global Parameters** |  |  |
-| global.local | Set to `true` if installing to a local K8s cluster (e.g.: Kind, Rancher Desktop, etc.) | false |
-| global.ingressClass | [IngressClass](https://kubernetes.io/zh-cn/docs/concepts/services-networking/ingress/#ingress-class) which is used to filter Ingress resources Higress Controller watches.<br />If there are multiple gateway instances deployed in the cluster, this parameter can be used to distinguish the scope of each gateway instance.<br />There are some special cases for special IngressClass values:<br />1. If set to "nginx", Higress Controller will watch Ingress resources with the `nginx` IngressClass or without any Ingress class.<br />2. If set to empty, Higress Controller will watch all Ingress resources in the K8s cluster. | higress |
-| global.watchNamespace | If not empty, Higress Controller will only watch resources in the specified namespace. When isolating different business systems using K8s namespace, if each namespace requires a standalone gateway instance, this parameter can be used to confine the Ingress watching of Higress within the given namespace. | "" |
-| global.disableAlpnH2 | Whether to disable HTTP/2 in ALPN | true |
-| global.enableStatus | If `true`, Higress Controller will update the `status` field of Ingress resources.<br />When migrating from Nginx Ingress, in order to avoid `status` field of Ingress objects being overwritten, this parameter needs to be set to false, so Higress won't write the entry IP to the `status` field of the corresponding Ingress object. | true |
-| global.enableIstioAPI | If `true`, Higress Controller will monitor istio resources as well | false |
-| global.enableGatewayAPI | If `true`, Higress Controller will monitor Gateway API resources as well | false |
-| global.istioNamespace | The namespace istio is installed to | istio-system |
-| **Core Paramters** |  |  |
-| higress-core.gateway.replicas | Number of Higress Gateway pods | 2 |
-| higress-core.controller.replicas | Number of Higress Controller pods | 1 |
-| **Console Paramters** |  |  |
-| higress-console.replicaCount | Number of Higress Console pods | 1 |
-| higress-console.service.type | K8s service type used by Higress Console | ClusterIP |
-| higress-console.domain | Domain used to access Higress Console | console.higress.io |
-| higress-console.tlsSecretName | Name of Secret resource used by TLS connections. | "" |
-| higress-console.web.login.prompt | Prompt message to be displayed on the login page | "" |
-| higress-console.admin.password.value | If not empty, the admin password will be configured to the specified value. | "" |
-| higress-console.admin.password.length | The length of random admin password generated during installation. Only works when `higress-console.admin.password.value` is not set. | 8 |
-| higress-console.o11y.enabled | If `true`, o11y suite (Grafana + Promethues) will be installed. | false |
-| higress-console.pvc.rwxSupported | Set to `false` when installing to a standard K8s cluster and the target cluster doesn't support the ReadWriteMany access mode of PersistentVolumeClaim. | true |
+## Higress for Kubernetes
+
+Higress is a cloud-native api gateway based on Alibaba's internal gateway practices.
+
+Powered by Istio and Envoy, Higress realizes the integration of the triple gateway architecture of traffic gateway, microservice gateway and security gateway, thereby greatly reducing the costs of deployment, operation and maintenance.
+
+## Setup Repo Info
+
+```console
+helm repo add higress.io https://higress.io/helm-charts
+helm repo update
+```
+
+## Install
+
+To install the chart with the release name `higress`:
+
+```console
+helm install higress -n higress-system higress.io/higress --create-namespace --render-subchart-notes
+```
+
+## Uninstall
+
+To uninstall/delete the higress deployment:
+
+```console
+helm delete higress -n higress-system
+```
+
+The command removes all the Kubernetes components associated with the chart and deletes the release.
+
+## Parameters
+
+## Values
+
+| Key | Type | Default | Description |
+|-----|------|---------|-------------|
+| clusterName | string | `""` |  |
+| controller.affinity | object | `{}` |  |
+| controller.automaticHttps.email | string | `""` |  |
+| controller.automaticHttps.enabled | bool | `true` |  |
+| controller.autoscaling.enabled | bool | `false` |  |
+| controller.autoscaling.maxReplicas | int | `5` |  |
+| controller.autoscaling.minReplicas | int | `1` |  |
+| controller.autoscaling.targetCPUUtilizationPercentage | int | `80` |  |
+| controller.env | object | `{}` |  |
+| controller.hub | string | `"higress-registry.cn-hangzhou.cr.aliyuncs.com/higress"` |  |
+| controller.image | string | `"higress"` |  |
+| controller.imagePullSecrets | list | `[]` |  |
+| controller.labels | object | `{}` |  |
+| controller.name | string | `"higress-controller"` |  |
+| controller.nodeSelector | object | `{}` |  |
+| controller.podAnnotations | object | `{}` |  |
+| controller.podSecurityContext | object | `{}` |  |
+| controller.ports[0].name | string | `"http"` |  |
+| controller.ports[0].port | int | `8888` |  |
+| controller.ports[0].protocol | string | `"TCP"` |  |
+| controller.ports[0].targetPort | int | `8888` |  |
+| controller.ports[1].name | string | `"http-solver"` |  |
+| controller.ports[1].port | int | `8889` |  |
+| controller.ports[1].protocol | string | `"TCP"` |  |
+| controller.ports[1].targetPort | int | `8889` |  |
+| controller.ports[2].name | string | `"grpc"` |  |
+| controller.ports[2].port | int | `15051` |  |
+| controller.ports[2].protocol | string | `"TCP"` |  |
+| controller.ports[2].targetPort | int | `15051` |  |
+| controller.probe.httpGet.path | string | `"/ready"` |  |
+| controller.probe.httpGet.port | int | `8888` |  |
+| controller.probe.initialDelaySeconds | int | `1` |  |
+| controller.probe.periodSeconds | int | `3` |  |
+| controller.probe.timeoutSeconds | int | `5` |  |
+| controller.rbac.create | bool | `true` |  |
+| controller.replicas | int | `1` | Number of Higress Controller pods |
+| controller.resources.limits.cpu | string | `"1000m"` |  |
+| controller.resources.limits.memory | string | `"2048Mi"` |  |
+| controller.resources.requests.cpu | string | `"500m"` |  |
+| controller.resources.requests.memory | string | `"2048Mi"` |  |
+| controller.securityContext | object | `{}` |  |
+| controller.service.type | string | `"ClusterIP"` |  |
+| controller.serviceAccount.annotations | object | `{}` | Annotations to add to the service account |
+| controller.serviceAccount.create | bool | `true` | Specifies whether a service account should be created |
+| controller.serviceAccount.name | string | `""` | If not set and create is true, a name is generated using the fullname template |
+| controller.tag | string | `""` |  |
+| controller.tolerations | list | `[]` |  |
+| downstream | object | `{"connectionBufferLimits":32768,"http2":{"initialConnectionWindowSize":1048576,"initialStreamWindowSize":65535,"maxConcurrentStreams":100},"idleTimeout":180,"maxRequestHeadersKb":60,"routeTimeout":0}` | Downstream config settings |
+| gateway.affinity | object | `{}` |  |
+| gateway.annotations | object | `{}` | Annotations to apply to all resources |
+| gateway.autoscaling.enabled | bool | `false` |  |
+| gateway.autoscaling.maxReplicas | int | `5` |  |
+| gateway.autoscaling.minReplicas | int | `1` |  |
+| gateway.autoscaling.targetCPUUtilizationPercentage | int | `80` |  |
+| gateway.containerSecurityContext | string | `nil` |  |
+| gateway.env | object | `{}` | Pod environment variables |
+| gateway.hostNetwork | bool | `false` |  |
+| gateway.httpPort | int | `80` |  |
+| gateway.httpsPort | int | `443` |  |
+| gateway.hub | string | `"higress-registry.cn-hangzhou.cr.aliyuncs.com/higress"` |  |
+| gateway.image | string | `"gateway"` |  |
+| gateway.kind | string | `"Deployment"` | Use a `DaemonSet` or `Deployment` |
+| gateway.labels | object | `{}` | Labels to apply to all resources |
+| gateway.metrics.enabled | bool | `false` | If true, create PodMonitor or VMPodScrape for gateway |
+| gateway.metrics.honorLabels | bool | `false` |  |
+| gateway.metrics.interval | string | `""` |  |
+| gateway.metrics.metricRelabelConfigs | list | `[]` | for operator.victoriametrics.com/v1beta1.VMPodScrape |
+| gateway.metrics.metricRelabelings | list | `[]` | for monitoring.coreos.com/v1.PodMonitor |
+| gateway.metrics.provider | string | `"monitoring.coreos.com"` | provider group name for CustomResourceDefinition, can be monitoring.coreos.com or operator.victoriametrics.com |
+| gateway.metrics.rawSpec | object | `{}` | some more raw podMetricsEndpoints spec |
+| gateway.metrics.relabelConfigs | list | `[]` |  |
+| gateway.metrics.relabelings | list | `[]` |  |
+| gateway.metrics.scrapeTimeout | string | `""` |  |
+| gateway.name | string | `"higress-gateway"` |  |
+| gateway.networkGateway | string | `""` | If specified, the gateway will act as a network gateway for the given network. |
+| gateway.nodeSelector | object | `{}` |  |
+| gateway.podAnnotations."prometheus.io/path" | string | `"/stats/prometheus"` |  |
+| gateway.podAnnotations."prometheus.io/port" | string | `"15020"` |  |
+| gateway.podAnnotations."prometheus.io/scrape" | string | `"true"` |  |
+| gateway.podAnnotations."sidecar.istio.io/inject" | string | `"false"` |  |
+| gateway.rbac.enabled | bool | `true` | If enabled, roles will be created to enable accessing certificates from Gateways. This is not needed when using http://gateway-api.org/. |
+| gateway.readinessFailureThreshold | int | `30` | The number of successive failed probes before indicating readiness failure. |
+| gateway.readinessInitialDelaySeconds | int | `1` | The initial delay for readiness probes in seconds. |
+| gateway.readinessPeriodSeconds | int | `2` | The period between readiness probes. |
+| gateway.readinessSuccessThreshold | int | `1` | The number of successive successed probes before indicating readiness success. |
+| gateway.readinessTimeoutSeconds | int | `3` | The readiness timeout seconds |
+| gateway.replicas | int | `2` | Number of Higress Gateway pods |
+| gateway.resources.limits.cpu | string | `"2000m"` |  |
+| gateway.resources.limits.memory | string | `"2048Mi"` |  |
+| gateway.resources.requests.cpu | string | `"2000m"` |  |
+| gateway.resources.requests.memory | string | `"2048Mi"` |  |
+| gateway.revision | string | `""` | revision declares which revision this gateway is a part of |
+| gateway.rollingMaxSurge | string | `"100%"` |  |
+| gateway.rollingMaxUnavailable | string | `"25%"` |  |
+| gateway.securityContext | string | `nil` | Define the security context for the pod. If unset, this will be automatically set to the minimum privileges required to bind to port 80 and 443. On Kubernetes 1.22+, this only requires the `net.ipv4.ip_unprivileged_port_start` sysctl. |
+| gateway.service.annotations | object | `{}` |  |
+| gateway.service.externalTrafficPolicy | string | `""` |  |
+| gateway.service.loadBalancerClass | string | `""` |  |
+| gateway.service.loadBalancerIP | string | `""` |  |
+| gateway.service.loadBalancerSourceRanges | list | `[]` |  |
+| gateway.service.ports[0].name | string | `"http2"` |  |
+| gateway.service.ports[0].port | int | `80` |  |
+| gateway.service.ports[0].protocol | string | `"TCP"` |  |
+| gateway.service.ports[0].targetPort | int | `80` |  |
+| gateway.service.ports[1].name | string | `"https"` |  |
+| gateway.service.ports[1].port | int | `443` |  |
+| gateway.service.ports[1].protocol | string | `"TCP"` |  |
+| gateway.service.ports[1].targetPort | int | `443` |  |
+| gateway.service.type | string | `"LoadBalancer"` | Type of service. Set to "None" to disable the service entirely |
+| gateway.serviceAccount.annotations | object | `{}` | Annotations to add to the service account |
+| gateway.serviceAccount.create | bool | `true` | If set, a service account will be created. Otherwise, the default is used |
+| gateway.serviceAccount.name | string | `""` | The name of the service account to use. If not set, the release name is used |
+| gateway.tag | string | `""` |  |
+| gateway.tolerations | list | `[]` |  |
+| gateway.unprivilegedPortSupported | string | `nil` |  |
+| global.autoscalingv2API | bool | `true` | whether to use autoscaling/v2 template for HPA settings for internal usage only, not to be configured by users. |
+| global.caAddress | string | `""` | The customized CA address to retrieve certificates for the pods in the cluster. CSR clients such as the Istio Agent and ingress gateways can use this to specify the CA endpoint. If not set explicitly, default to the Istio discovery address. |
+| global.caName | string | `""` | The name of the CA for workload certificates. For example, when caName=GkeWorkloadCertificate, GKE workload certificates will be used as the certificates for workloads. The default value is "" and when caName="", the CA will be configured by other mechanisms (e.g., environmental variable CA_PROVIDER). |
+| global.configCluster | bool | `false` | Configure a remote cluster as the config cluster for an external istiod. |
+| global.defaultPodDisruptionBudget | object | `{"enabled":false}` | enable pod disruption budget for the control plane, which is used to ensure Istio control plane components are gradually upgraded or recovered. |
+| global.defaultResources | object | `{"requests":{"cpu":"10m"}}` | A minimal set of requested resources to applied to all deployments so that Horizontal Pod Autoscaler will be able to function (if set). Each component can overwrite these default values by adding its own resources block in the relevant section below and setting the desired resources values. |
+| global.defaultUpstreamConcurrencyThreshold | int | `10000` |  |
+| global.disableAlpnH2 | bool | `false` | Whether to disable HTTP/2 in ALPN |
+| global.enableGatewayAPI | bool | `false` | If true, Higress Controller will monitor Gateway API resources as well |
+| global.enableH3 | bool | `false` |  |
+| global.enableIPv6 | bool | `false` |  |
+| global.enableIstioAPI | bool | `true` | If true, Higress Controller will monitor istio resources as well |
+| global.enableLDSCache | bool | `true` |  |
+| global.enableProxyProtocol | bool | `false` |  |
+| global.enablePushAllMCPClusters | bool | `true` |  |
+| global.enableSRDS | bool | `true` |  |
+| global.enableStatus | bool | `true` | If true, Higress Controller will update the status field of Ingress resources. When migrating from Nginx Ingress, in order to avoid status field of Ingress objects being overwritten, this parameter needs to be set to false, so Higress won't write the entry IP to the status field of the corresponding Ingress object. |
+| global.externalIstiod | bool | `false` | Configure a remote cluster data plane controlled by an external istiod. When set to true, istiod is not deployed locally and only a subset of the other discovery charts are enabled. |
+| global.hostRDSMergeSubset | bool | `false` |  |
+| global.hub | string | `"higress-registry.cn-hangzhou.cr.aliyuncs.com/higress"` | Default hub for Istio images. Releases are published to docker hub under 'istio' project. Dev builds from prow are on gcr.io |
+| global.imagePullPolicy | string | `""` | Specify image pull policy if default behavior isn't desired. Default behavior: latest images will be Always else IfNotPresent. |
+| global.imagePullSecrets | list | `[]` | ImagePullSecrets for all ServiceAccount, list of secrets in the same namespace to use for pulling any images in pods that reference this ServiceAccount. For components that don't use ServiceAccounts (i.e. grafana, servicegraph, tracing) ImagePullSecrets will be added to the corresponding Deployment(StatefulSet) objects. Must be set for any cluster configured with private docker registry. |
+| global.ingressClass | string | `"higress"` | IngressClass filters which ingress resources the higress controller watches. The default ingress class is higress. There are some special cases for special ingress class. 1. When the ingress class is set as nginx, the higress controller will watch ingress resources with the nginx ingress class or without any ingress class. 2. When the ingress class is set empty, the higress controller will watch all ingress resources in the k8s cluster. |
+| global.istioNamespace | string | `"istio-system"` | Used to locate istiod. |
+| global.istiod | object | `{"enableAnalysis":false}` | Enabled by default in master for maximising testing. |
+| global.jwtPolicy | string | `"third-party-jwt"` | Configure the policy for validating JWT. Currently, two options are supported: "third-party-jwt" and "first-party-jwt". |
+| global.kind | bool | `false` |  |
+| global.liteMetrics | bool | `false` |  |
+| global.local | bool | `false` | When deploying to a local cluster (e.g.: kind cluster), set this to true. |
+| global.logAsJson | bool | `false` |  |
+| global.logging | object | `{"level":"default:info"}` | Comma-separated minimum per-scope logging level of messages to output, in the form of <scope>:<level>,<scope>:<level> The control plane has different scopes depending on component, but can configure default log level across all components If empty, default scope and level will be used as configured in code |
+| global.meshID | string | `""` | If the mesh admin does not specify a value, Istio will use the value of the mesh's Trust Domain. The best practice is to select a proper Trust Domain value. |
+| global.meshNetworks | object | `{}` |  |
+| global.mountMtlsCerts | bool | `false` | Use the user-specified, secret volume mounted key and certs for Pilot and workloads. |
+| global.multiCluster.clusterName | string | `""` | Should be set to the name of the cluster this installation will run in. This is required for sidecar injection to properly label proxies |
+| global.multiCluster.enabled | bool | `true` | Set to true to connect two kubernetes clusters via their respective ingressgateway services when pods in each cluster cannot directly talk to one another. All clusters should be using Istio mTLS and must have a shared root CA for this model to work. |
+| global.network | string | `""` | Network defines the network this cluster belong to. This name corresponds to the networks in the map of mesh networks. |
+| global.o11y | object | `{"enabled":false,"promtail":{"image":{"repository":"higress-registry.cn-hangzhou.cr.aliyuncs.com/higress/promtail","tag":"2.9.4"},"port":3101,"resources":{"limits":{"cpu":"500m","memory":"2Gi"}},"securityContext":{}}}` | Observability (o11y) configurations |
+| global.omitSidecarInjectorConfigMap | bool | `false` |  |
+| global.onDemandRDS | bool | `false` |  |
+| global.oneNamespace | bool | `false` | Whether to restrict the applications namespace the controller manages; If not set, controller watches all namespaces |
+| global.onlyPushRouteCluster | bool | `true` |  |
+| global.operatorManageWebhooks | bool | `false` | Configure whether Operator manages webhook configurations. The current behavior of Istiod is to manage its own webhook configurations. When this option is set as true, Istio Operator, instead of webhooks, manages the webhook configurations. When this option is set as false, webhooks manage their own webhook configurations. |
+| global.pilotCertProvider | string | `"istiod"` | Configure the certificate provider for control plane communication. Currently, two providers are supported: "kubernetes" and "istiod". As some platforms may not have kubernetes signing APIs, Istiod is the default |
+| global.priorityClassName | string | `""` | Kubernetes >=v1.11.0 will create two PriorityClass, including system-cluster-critical and system-node-critical, it is better to configure this in order to make sure your Istio pods will not be killed because of low priority class. Refer to https://kubernetes.io/docs/concepts/configuration/pod-priority-preemption/#priorityclass for more detail. |
+| global.proxy.autoInject | string | `"enabled"` | This controls the 'policy' in the sidecar injector. |
+| global.proxy.clusterDomain | string | `"cluster.local"` | CAUTION: It is important to ensure that all Istio helm charts specify the same clusterDomain value cluster domain. Default value is "cluster.local". |
+| global.proxy.componentLogLevel | string | `"misc:error"` | Per Component log level for proxy, applies to gateways and sidecars. If a component level is not set, then the global "logLevel" will be used. |
+| global.proxy.enableCoreDump | bool | `false` | If set, newly injected sidecars will have core dumps enabled. |
+| global.proxy.excludeIPRanges | string | `""` |  |
+| global.proxy.excludeInboundPorts | string | `""` |  |
+| global.proxy.excludeOutboundPorts | string | `""` |  |
+| global.proxy.holdApplicationUntilProxyStarts | bool | `false` | Controls if sidecar is injected at the front of the container list and blocks the start of the other containers until the proxy is ready |
+| global.proxy.image | string | `"proxyv2"` |  |
+| global.proxy.includeIPRanges | string | `"*"` | istio egress capture allowlist https://istio.io/docs/tasks/traffic-management/egress.html#calling-external-services-directly example: includeIPRanges: "172.30.0.0/16,172.20.0.0/16" would only capture egress traffic on those two IP Ranges, all other outbound traffic would be allowed by the sidecar |
+| global.proxy.includeInboundPorts | string | `"*"` |  |
+| global.proxy.includeOutboundPorts | string | `""` |  |
+| global.proxy.logLevel | string | `"warning"` | Log level for proxy, applies to gateways and sidecars. Expected values are: trace|debug|info|warning|error|critical|off |
+| global.proxy.privileged | bool | `false` | If set to true, istio-proxy container will have privileged securityContext |
+| global.proxy.readinessFailureThreshold | int | `30` | The number of successive failed probes before indicating readiness failure. |
+| global.proxy.readinessInitialDelaySeconds | int | `1` | The initial delay for readiness probes in seconds. |
+| global.proxy.readinessPeriodSeconds | int | `2` | The period between readiness probes. |
+| global.proxy.readinessSuccessThreshold | int | `30` | The number of successive successed probes before indicating readiness success. |
+| global.proxy.readinessTimeoutSeconds | int | `3` | The readiness timeout seconds |
+| global.proxy.resources | object | `{"limits":{"cpu":"2000m","memory":"1024Mi"},"requests":{"cpu":"100m","memory":"128Mi"}}` | Resources for the sidecar. |
+| global.proxy.statusPort | int | `15020` | Default port for Pilot agent health checks. A value of 0 will disable health checking. |
+| global.proxy.tracer | string | `""` | Specify which tracer to use. One of: lightstep, datadog, stackdriver. If using stackdriver tracer outside GCP, set env GOOGLE_APPLICATION_CREDENTIALS to the GCP credential file. |
+| global.proxy_init.image | string | `"proxyv2"` | Base name for the proxy_init container, used to configure iptables. |
+| global.proxy_init.resources.limits.cpu | string | `"2000m"` |  |
+| global.proxy_init.resources.limits.memory | string | `"1024Mi"` |  |
+| global.proxy_init.resources.requests.cpu | string | `"10m"` |  |
+| global.proxy_init.resources.requests.memory | string | `"10Mi"` |  |
+| global.remotePilotAddress | string | `""` | configure remote pilot and istiod service and endpoint |
+| global.sds.token | object | `{"aud":"istio-ca"}` | The JWT token for SDS and the aud field of such JWT. See RFC 7519, section 4.1.3. When a CSR is sent from Istio Agent to the CA (e.g. Istiod), this aud is to make sure the JWT is intended for the CA. |
+| global.sts.servicePort | int | `0` | The service port used by Security Token Service (STS) server to handle token exchange requests. Setting this port to a non-zero value enables STS server. |
+| global.tracer | object | `{"datadog":{"address":"$(HOST_IP):8126"},"lightstep":{"accessToken":"","address":""},"stackdriver":{"debug":false,"maxNumberOfAnnotations":200,"maxNumberOfAttributes":200,"maxNumberOfMessageEvents":200}}` | Configuration for each of the supported tracers |
+| global.tracer.datadog | object | `{"address":"$(HOST_IP):8126"}` | Configuration for envoy to send trace data to LightStep. Disabled by default. address: the <host>:<port> of the satellite pool accessToken: required for sending data to the pool  |
+| global.tracer.datadog.address | string | `"$(HOST_IP):8126"` | Host:Port for submitting traces to the Datadog agent. |
+| global.tracer.lightstep.accessToken | string | `""` | example: abcdefg1234567 |
+| global.tracer.lightstep.address | string | `""` | example: lightstep-satellite:443 |
+| global.tracer.stackdriver.debug | bool | `false` | enables trace output to stdout. |
+| global.tracer.stackdriver.maxNumberOfAnnotations | int | `200` | The global default max number of annotation events per span. |
+| global.tracer.stackdriver.maxNumberOfAttributes | int | `200` | The global default max number of attributes per span. |
+| global.tracer.stackdriver.maxNumberOfMessageEvents | int | `200` | The global default max number of message events per span. |
+| global.useMCP | bool | `false` | Use the Mesh Control Protocol (MCP) for configuring Istiod. Requires an MCP source. |
+| global.watchNamespace | string | `""` | If not empty, Higress Controller will only watch resources in the specified namespace. When isolating different business systems using K8s namespace, if each namespace requires a standalone gateway instance, this parameter can be used to confine the Ingress watching of Higress within the given namespace. |
+| global.xdsMaxRecvMsgSize | string | `"104857600"` |  |
+| hub | string | `"higress-registry.cn-hangzhou.cr.aliyuncs.com/higress"` |  |
+| meshConfig | object | `{"enablePrometheusMerge":true,"rootNamespace":null,"trustDomain":"cluster.local"}` | meshConfig defines runtime configuration of components, including Istiod and istio-agent behavior See https://istio.io/docs/reference/config/istio.mesh.v1alpha1/ for all available options |
+| meshConfig.rootNamespace | string | `nil` | The namespace to treat as the administrative root namespace for Istio configuration. When processing a leaf namespace Istio will search for declarations in that namespace first and if none are found it will search in the root namespace. Any matching declaration found in the root namespace is processed as if it were declared in the leaf namespace. |
+| meshConfig.trustDomain | string | `"cluster.local"` | The trust domain corresponds to the trust root of a system Refer to https://github.com/spiffe/spiffe/blob/master/standards/SPIFFE-ID.md#21-trust-domain |
+| pilot.autoscaleEnabled | bool | `false` |  |
+| pilot.autoscaleMax | int | `5` |  |
+| pilot.autoscaleMin | int | `1` |  |
+| pilot.configMap | bool | `true` | Install the mesh config map, generated from values.yaml. If false, pilot wil use default values (by default) or user-supplied values. |
+| pilot.configSource | object | `{"subscribedResources":[]}` | This is used to set the source of configuration for the associated address in configSource, if nothing is specified the default MCP is assumed. |
+| pilot.cpu.targetAverageUtilization | int | `80` |  |
+| pilot.deploymentLabels | object | `{}` | Additional labels to apply to the deployment. |
+| pilot.enableProtocolSniffingForInbound | bool | `true` | if protocol sniffing is enabled for inbound |
+| pilot.enableProtocolSniffingForOutbound | bool | `true` | if protocol sniffing is enabled for outbound |
+| pilot.env.PILOT_ENABLE_CROSS_CLUSTER_WORKLOAD_ENTRY | string | `"false"` |  |
+| pilot.env.PILOT_ENABLE_METADATA_EXCHANGE | string | `"false"` |  |
+| pilot.env.PILOT_SCOPE_GATEWAY_TO_NAMESPACE | string | `"false"` |  |
+| pilot.env.VALIDATION_ENABLED | string | `"false"` |  |
+| pilot.hub | string | `"higress-registry.cn-hangzhou.cr.aliyuncs.com/higress"` |  |
+| pilot.image | string | `"pilot"` | Can be a full hub/image:tag |
+| pilot.jwksResolverExtraRootCA | string | `""` | You can use jwksResolverExtraRootCA to provide a root certificate in PEM format. This will then be trusted by pilot when resolving JWKS URIs. |
+| pilot.keepaliveMaxServerConnectionAge | string | `"30m"` | The following is used to limit how long a sidecar can be connected to a pilot. It balances out load across pilot instances at the cost of increasing system churn. |
+| pilot.nodeSelector | object | `{}` |  |
+| pilot.plugins | list | `[]` |  |
+| pilot.podAnnotations | object | `{}` |  |
+| pilot.podLabels | object | `{}` | Additional labels to apply on the pod level for monitoring and logging configuration. |
+| pilot.replicaCount | int | `1` |  |
+| pilot.resources | object | `{"requests":{"cpu":"500m","memory":"2048Mi"}}` | Resources for a small pilot install |
+| pilot.rollingMaxSurge | string | `"100%"` |  |
+| pilot.rollingMaxUnavailable | string | `"25%"` |  |
+| pilot.serviceAnnotations | object | `{}` |  |
+| pilot.tag | string | `""` |  |
+| pilot.traceSampling | float | `1` |  |
+| revision | string | `""` |  |
+| tracing.enable | bool | `false` |  |
+| tracing.sampling | int | `100` |  |
+| tracing.skywalking.port | int | `11800` |  |
+| tracing.skywalking.service | string | `""` |  |
+| tracing.timeout | int | `500` |  |
+| upstream | object | `{"connectionBufferLimits":10485760,"idleTimeout":10}` | Upstream config settings |
--- a/helm/higress/README.md.gotmpl
+++ b/helm/higress/README.md.gotmpl
@@ -0,0 +1,34 @@
+## Higress for Kubernetes
+
+Higress is a cloud-native api gateway based on Alibaba's internal gateway practices.
+
+Powered by Istio and Envoy, Higress realizes the integration of the triple gateway architecture of traffic gateway, microservice gateway and security gateway, thereby greatly reducing the costs of deployment, operation and maintenance.
+
+## Setup Repo Info
+
+```console
+helm repo add higress.io https://higress.io/helm-charts
+helm repo update
+```
+
+## Install
+
+To install the chart with the release name `higress`:
+
+```console
+helm install higress -n higress-system higress.io/higress --create-namespace --render-subchart-notes
+```
+
+## Uninstall
+
+To uninstall/delete the higress deployment:
+
+```console
+helm delete higress -n higress-system
+```
+
+The command removes all the Kubernetes components associated with the chart and deletes the release.
+
+## Parameters
+
+{{ template "chart.valuesSection" . }}
--- a/istio/istio
+++ b/istio/istio
--- a/istio/proxy
+++ b/istio/proxy
--- a/pkg/bootstrap/server.go
+++ b/pkg/bootstrap/server.go
@@ -41,11 +41,11 @@ import (
 	"istio.io/istio/pkg/config/schema/kind"
 	"istio.io/istio/pkg/keepalive"
 	istiokube "istio.io/istio/pkg/kube"
+	"istio.io/istio/pkg/log"
 	"istio.io/istio/pkg/security"
 	"istio.io/istio/security/pkg/server/ca/authenticate"
 	"istio.io/istio/security/pkg/server/ca/authenticate/kubeauth"
 	"istio.io/pkg/ledger"
-	"istio.io/pkg/log"
 	"k8s.io/client-go/rest"
 	"k8s.io/client-go/tools/cache"

--- a/pkg/cert/certmgr.go
+++ b/pkg/cert/certmgr.go
@@ -173,7 +173,7 @@ func (s *CertMgr) Reconcile(ctx context.Context, oldConfig *Config, newConfig *C
 		s.cache.Start()
 		// sync domains
 		s.configMgr.SetConfig(newConfig)
-		CertLog.Infof("certMgr start to manageSync domains:+v%", newDomains)
+		CertLog.Infof("certMgr start to manageSync domains: %+v", newDomains)
 		s.manageSync(context.Background(), newDomains)
 		CertLog.Infof("certMgr manageSync domains done")
 	} else {
--- a/pkg/cert/log.go
+++ b/pkg/cert/log.go
@@ -14,6 +14,6 @@

 package cert

-import "istio.io/pkg/log"
+import "istio.io/istio/pkg/log"

-var CertLog = log.RegisterScope("cert", "Higress Cert process.", 0)
+var CertLog = log.RegisterScope("cert", "Higress Cert process.")
--- a/pkg/cmd/server.go
+++ b/pkg/cmd/server.go
@@ -25,7 +25,7 @@ import (
 	"istio.io/istio/pkg/config/constants"
 	"istio.io/istio/pkg/env"
 	"istio.io/istio/pkg/keepalive"
-	"istio.io/pkg/log"
+	"istio.io/istio/pkg/log"
 )

 var (
--- a/pkg/ingress/config/ingress_config.go
+++ b/pkg/ingress/config/ingress_config.go
@@ -303,21 +303,21 @@ func (m *IngressConfig) listFromIngressControllers(typ config.GroupVersionKind,
 	common.SortIngressByCreationTime(configs)
 	wrapperConfigs := m.createWrapperConfigs(configs)

-	IngressLog.Infof("resource type %s, configs number %d", typ, len(wrapperConfigs))
+	var result []config.Config
 	switch typ {
 	case gvk.Gateway:
-		return m.convertGateways(wrapperConfigs)
+		result = m.convertGateways(wrapperConfigs)
 	case gvk.VirtualService:
-		return m.convertVirtualService(wrapperConfigs)
+		result = m.convertVirtualService(wrapperConfigs)
 	case gvk.DestinationRule:
-		return m.convertDestinationRule(wrapperConfigs)
+		result = m.convertDestinationRule(wrapperConfigs)
 	case gvk.ServiceEntry:
-		return m.convertServiceEntry(wrapperConfigs)
+		result = m.convertServiceEntry(wrapperConfigs)
 	case gvk.WasmPlugin:
-		return m.convertWasmPlugin(wrapperConfigs)
+		result = m.convertWasmPlugin(wrapperConfigs)
 	}
-
-	return nil
+	IngressLog.Infof("resource type %s, ingress number %d, convert configs number %d", typ, len(configs), len(result))
+	return result
 }

 func (m *IngressConfig) listFromGatewayControllers(typ config.GroupVersionKind, namespace string) []config.Config {
@@ -712,7 +712,6 @@ func (m *IngressConfig) convertDestinationRule(configs []common.WrapperConfig) [

 	if m.RegistryReconciler != nil {
 		drws := m.RegistryReconciler.GetAllDestinationRuleWrapper()
-		IngressLog.Infof("Found mcp destinationRules: %v", drws)
 		for _, destinationRuleWrapper := range drws {
 			serviceName := destinationRuleWrapper.ServiceKey.ServiceFQDN
 			dr, exist := destinationRules[serviceName]
@@ -882,7 +881,7 @@ func (m *IngressConfig) convertIstioWasmPlugin(obj *higressext.WasmPlugin) (*ext
 	if result.PluginConfig != nil {
 		return result, nil
 	}
-	if !obj.DefaultConfigDisable {
+	if !isBoolValueTrue(obj.DefaultConfigDisable) {
 		result.PluginConfig = obj.DefaultConfig
 	}
 	hasValidRule := false
@@ -894,7 +893,7 @@ func (m *IngressConfig) convertIstioWasmPlugin(obj *higressext.WasmPlugin) (*ext
 		}
 		var ruleValues []*_struct.Value
 		for _, rule := range obj.MatchRules {
-			if rule.ConfigDisable {
+			if isBoolValueTrue(rule.ConfigDisable) {
 				continue
 			}
 			if rule.Config == nil {
@@ -906,6 +905,7 @@ func (m *IngressConfig) convertIstioWasmPlugin(obj *higressext.WasmPlugin) (*ext
 				StructValue: rule.Config,
 			}

+			validRule := false
 			var matchItems []*_struct.Value
 			// match ingress
 			for _, ing := range rule.Ingress {
@@ -916,6 +916,7 @@ func (m *IngressConfig) convertIstioWasmPlugin(obj *higressext.WasmPlugin) (*ext
 				})
 			}
 			if len(matchItems) > 0 {
+				validRule = true
 				v.StructValue.Fields["_match_route_"] = &_struct.Value{
 					Kind: &_struct.Value_ListValue{
 						ListValue: &_struct.ListValue{
@@ -923,12 +924,9 @@ func (m *IngressConfig) convertIstioWasmPlugin(obj *higressext.WasmPlugin) (*ext
 						},
 					},
 				}
-				ruleValues = append(ruleValues, &_struct.Value{
-					Kind: v,
-				})
-				continue
 			}
 			// match service
+			matchItems = nil
 			for _, service := range rule.Service {
 				matchItems = append(matchItems, &_struct.Value{
 					Kind: &_struct.Value_StringValue{
@@ -937,6 +935,7 @@ func (m *IngressConfig) convertIstioWasmPlugin(obj *higressext.WasmPlugin) (*ext
 				})
 			}
 			if len(matchItems) > 0 {
+				validRule = true
 				v.StructValue.Fields["_match_service_"] = &_struct.Value{
 					Kind: &_struct.Value_ListValue{
 						ListValue: &_struct.ListValue{
@@ -944,12 +943,9 @@ func (m *IngressConfig) convertIstioWasmPlugin(obj *higressext.WasmPlugin) (*ext
 						},
 					},
 				}
-				ruleValues = append(ruleValues, &_struct.Value{
-					Kind: v,
-				})
-				continue
 			}
 			// match domain
+			matchItems = nil
 			for _, domain := range rule.Domain {
 				matchItems = append(matchItems, &_struct.Value{
 					Kind: &_struct.Value_StringValue{
@@ -957,19 +953,23 @@ func (m *IngressConfig) convertIstioWasmPlugin(obj *higressext.WasmPlugin) (*ext
 					},
 				})
 			}
-			if len(matchItems) == 0 {
+			if len(matchItems) > 0 {
+				validRule = true
+				v.StructValue.Fields["_match_domain_"] = &_struct.Value{
+					Kind: &_struct.Value_ListValue{
+						ListValue: &_struct.ListValue{
+							Values: matchItems,
+						},
+					},
+				}
+			}
+			if validRule {
+				ruleValues = append(ruleValues, &_struct.Value{
+					Kind: v,
+				})
+			} else {
 				return nil, fmt.Errorf("invalid match rule has no match condition, rule:%v", rule)
 			}
-			v.StructValue.Fields["_match_domain_"] = &_struct.Value{
-				Kind: &_struct.Value_ListValue{
-					ListValue: &_struct.ListValue{
-						Values: matchItems,
-					},
-				},
-			}
-			ruleValues = append(ruleValues, &_struct.Value{
-				Kind: v,
-			})
 		}
 		if len(ruleValues) > 0 {
 			hasValidRule = true
@@ -982,13 +982,17 @@ func (m *IngressConfig) convertIstioWasmPlugin(obj *higressext.WasmPlugin) (*ext
 			}
 		}
 	}
-	if !hasValidRule && obj.DefaultConfigDisable {
+	if !hasValidRule && isBoolValueTrue(obj.DefaultConfigDisable) {
 		return nil, nil
 	}
 	return result, nil

 }

+func isBoolValueTrue(b *wrappers.BoolValue) bool {
+	return b != nil && b.Value
+}
+
 func (m *IngressConfig) AddOrUpdateWasmPlugin(clusterNamespacedName util.ClusterNamespacedName) {
 	if clusterNamespacedName.Namespace != m.namespace {
 		return
--- a/pkg/ingress/config/kingress_config.go
+++ b/pkg/ingress/config/kingress_config.go
@@ -493,7 +493,7 @@ func (m *KIngressConfig) HasSynced() bool {
 	defer m.mutex.RUnlock()

 	for _, remoteIngressController := range m.remoteIngressControllers {
-		IngressLog.Info("In Kingress Synced.", remoteIngressController)
+		IngressLog.Info("In Kingress Synced.")
 		if !remoteIngressController.HasSynced() {
 			return false
 		}
--- a/pkg/ingress/kube/annotations/downstreamtls.go
+++ b/pkg/ingress/kube/annotations/downstreamtls.go
@@ -15,6 +15,7 @@
 package annotations

 import (
+	"fmt"
 	"strings"

 	networking "istio.io/api/networking/v1alpha3"
@@ -27,9 +28,11 @@ import (
 )

 const (
-	authTLSSecret      = "auth-tls-secret"
-	sslCipher          = "ssl-cipher"
-	gatewaySdsCaSuffix = "-cacert"
+	authTLSSecret           = "auth-tls-secret"
+	sslCipher               = "ssl-cipher"
+	gatewaySdsCaSuffix      = "-cacert"
+	annotationMinTLSVersion = "tls-min-protocol-version"
+	annotationMaxTLSVersion = "tls-max-protocol-version"
 )

 var (
@@ -41,6 +44,8 @@ type DownstreamTLSConfig struct {
 	CipherSuites []string
 	Mode         networking.ServerTLSSettings_TLSmode
 	CASecretName types.NamespacedName
+	MinVersion   string
+	MaxVersion   string
 }

 type downstreamTLS struct{}
@@ -82,6 +87,14 @@ func (d downstreamTLS) Parse(annotations Annotations, config *Ingress, _ *Global
 		downstreamTLSConfig.CipherSuites = validCipherSuite
 	}

+	if minVersion, err := annotations.ParseStringASAP(annotationMinTLSVersion); err == nil {
+		downstreamTLSConfig.MinVersion = minVersion
+	}
+
+	if maxVersion, err := annotations.ParseStringASAP(annotationMaxTLSVersion); err == nil {
+		downstreamTLSConfig.MaxVersion = maxVersion
+	}
+
 	return nil
 }

@@ -107,11 +120,44 @@ func (d downstreamTLS) ApplyGateway(gateway *networking.Gateway, config *Ingress
 			if len(downstreamTLSConfig.CipherSuites) != 0 {
 				server.Tls.CipherSuites = downstreamTLSConfig.CipherSuites
 			}
+
+			if downstreamTLSConfig.MinVersion != "" {
+				if version, err := convertTLSVersion(downstreamTLSConfig.MinVersion); err != nil {
+					IngressLog.Errorf("Invalid minimum TLS version: %v", err)
+				} else {
+					server.Tls.MinProtocolVersion = version
+				}
+			}
+
+			if downstreamTLSConfig.MaxVersion != "" {
+				if version, err := convertTLSVersion(downstreamTLSConfig.MaxVersion); err != nil {
+					IngressLog.Errorf("Invalid maximum TLS version: %v", err)
+				} else {
+					server.Tls.MaxProtocolVersion = version
+				}
+			}
+
 		}
 	}
 }

 func needDownstreamTLS(annotations Annotations) bool {
 	return annotations.HasASAP(sslCipher) ||
-		annotations.HasASAP(authTLSSecret)
+		annotations.HasASAP(authTLSSecret) ||
+		annotations.HasASAP(annotationMinTLSVersion) ||
+		annotations.HasASAP(annotationMaxTLSVersion)
+}
+
+func convertTLSVersion(version string) (networking.ServerTLSSettings_TLSProtocol, error) {
+	switch version {
+	case "TLSv1.0":
+		return networking.ServerTLSSettings_TLSV1_0, nil
+	case "TLSv1.1":
+		return networking.ServerTLSSettings_TLSV1_1, nil
+	case "TLSv1.2":
+		return networking.ServerTLSSettings_TLSV1_2, nil
+	case "TLSv1.3":
+		return networking.ServerTLSSettings_TLSV1_3, nil
+	}
+	return networking.ServerTLSSettings_TLS_AUTO, fmt.Errorf("invalid TLS version: %s. Valid values are: TLSv1.0, TLSv1.1, TLSv1.2, TLSv1.3", version)
 }
--- a/pkg/ingress/kube/annotations/downstreamtls_test.go
+++ b/pkg/ingress/kube/annotations/downstreamtls_test.go
@@ -26,11 +26,15 @@ var parser = downstreamTLS{}

 func TestParse(t *testing.T) {
 	testCases := []struct {
+		name   string
 		input  map[string]string
 		expect *DownstreamTLSConfig
 	}{
-		{},
 		{
+			name: "empty config",
+		},
+		{
+			name: "ssl cipher only",
 			input: map[string]string{
 				buildNginxAnnotationKey(sslCipher): "ECDHE-RSA-AES256-GCM-SHA384:AES128-SHA",
 			},
@@ -40,9 +44,24 @@ func TestParse(t *testing.T) {
 			},
 		},
 		{
+			name: "with TLS version config",
 			input: map[string]string{
-				buildNginxAnnotationKey(authTLSSecret): "test",
-				buildNginxAnnotationKey(sslCipher):     "ECDHE-RSA-AES256-GCM-SHA384:AES128-SHA",
+				buildNginxAnnotationKey(annotationMinTLSVersion): "TLSv1.2",
+				buildNginxAnnotationKey(annotationMaxTLSVersion): "TLSv1.3",
+			},
+			expect: &DownstreamTLSConfig{
+				Mode:       networking.ServerTLSSettings_SIMPLE,
+				MinVersion: "TLSv1.2",
+				MaxVersion: "TLSv1.3",
+			},
+		},
+		{
+			name: "complete config",
+			input: map[string]string{
+				buildNginxAnnotationKey(authTLSSecret):           "test",
+				buildNginxAnnotationKey(sslCipher):               "ECDHE-RSA-AES256-GCM-SHA384:AES128-SHA",
+				buildNginxAnnotationKey(annotationMinTLSVersion): "TLSv1.2",
+				buildNginxAnnotationKey(annotationMaxTLSVersion): "TLSv1.3",
 			},
 			expect: &DownstreamTLSConfig{
 				CASecretName: types.NamespacedName{
@@ -51,34 +70,79 @@ func TestParse(t *testing.T) {
 				},
 				Mode:         networking.ServerTLSSettings_MUTUAL,
 				CipherSuites: []string{"ECDHE-RSA-AES256-GCM-SHA384", "AES128-SHA"},
-			},
-		},
-		{
-			input: map[string]string{
-				buildHigressAnnotationKey(authTLSSecret):   "test/foo",
-				DefaultAnnotationsPrefix + "/" + sslCipher: "ECDHE-RSA-AES256-GCM-SHA384:AES128-SHA",
-			},
-			expect: &DownstreamTLSConfig{
-				CASecretName: types.NamespacedName{
-					Namespace: "test",
-					Name:      "foo",
-				},
-				Mode:         networking.ServerTLSSettings_MUTUAL,
-				CipherSuites: []string{"ECDHE-RSA-AES256-GCM-SHA384", "AES128-SHA"},
+				MinVersion:   "TLSv1.2",
+				MaxVersion:   "TLSv1.3",
 			},
 		},
 	}

-	for _, testCase := range testCases {
-		t.Run("", func(t *testing.T) {
+	for _, tc := range testCases {
+		t.Run(tc.name, func(t *testing.T) {
 			config := &Ingress{
 				Meta: Meta{
 					Namespace: "foo",
 				},
 			}
-			_ = parser.Parse(testCase.input, config, nil)
-			if !reflect.DeepEqual(testCase.expect, config.DownstreamTLS) {
-				t.Fatalf("Should be equal")
+			err := parser.Parse(tc.input, config, nil)
+			if err != nil {
+				t.Fatalf("Parse failed: %v", err)
+			}
+			if !reflect.DeepEqual(tc.expect, config.DownstreamTLS) {
+				t.Fatalf("Parse result mismatch:\nExpect: %+v\nGot: %+v", tc.expect, config.DownstreamTLS)
+			}
+		})
+	}
+}
+
+func TestConvertTLSVersion(t *testing.T) {
+	testCases := []struct {
+		name    string
+		version string
+		expect  networking.ServerTLSSettings_TLSProtocol
+		wantErr bool
+	}{
+		{
+			name:    "TLS 1.0",
+			version: "TLSv1.0",
+			expect:  networking.ServerTLSSettings_TLSV1_0,
+		},
+		{
+			name:    "TLS 1.1",
+			version: "TLSv1.1",
+			expect:  networking.ServerTLSSettings_TLSV1_1,
+		},
+		{
+			name:    "TLS 1.2",
+			version: "TLSv1.2",
+			expect:  networking.ServerTLSSettings_TLSV1_2,
+		},
+		{
+			name:    "TLS 1.3",
+			version: "TLSv1.3",
+			expect:  networking.ServerTLSSettings_TLSV1_3,
+		},
+		{
+			name:    "invalid version",
+			version: "invalid",
+			expect:  networking.ServerTLSSettings_TLS_AUTO,
+			wantErr: true,
+		},
+	}
+
+	for _, tc := range testCases {
+		t.Run(tc.name, func(t *testing.T) {
+			result, err := convertTLSVersion(tc.version)
+			if tc.wantErr {
+				if err == nil {
+					t.Error("Expected error but got none")
+				}
+			} else {
+				if err != nil {
+					t.Errorf("Unexpected error: %v", err)
+				}
+				if result != tc.expect {
+					t.Errorf("Expected %v but got %v", tc.expect, result)
+				}
 			}
 		})
 	}
@@ -86,11 +150,13 @@ func TestParse(t *testing.T) {

 func TestApplyGateway(t *testing.T) {
 	testCases := []struct {
+		name   string
 		input  *networking.Gateway
 		config *Ingress
 		expect *networking.Gateway
 	}{
 		{
+			name: "apply TLS version",
 			input: &networking.Gateway{
 				Servers: []*networking.Server{
 					{
@@ -105,7 +171,8 @@ func TestApplyGateway(t *testing.T) {
 			},
 			config: &Ingress{
 				DownstreamTLS: &DownstreamTLSConfig{
-					CipherSuites: []string{"ECDHE-RSA-AES256-GCM-SHA384"},
+					MinVersion: "TLSv1.2",
+					MaxVersion: "TLSv1.3",
 				},
 			},
 			expect: &networking.Gateway{
@@ -115,14 +182,16 @@ func TestApplyGateway(t *testing.T) {
 							Protocol: "HTTPS",
 						},
 						Tls: &networking.ServerTLSSettings{
-							Mode:         networking.ServerTLSSettings_SIMPLE,
-							CipherSuites: []string{"ECDHE-RSA-AES256-GCM-SHA384"},
+							Mode:               networking.ServerTLSSettings_SIMPLE,
+							MinProtocolVersion: networking.ServerTLSSettings_TLSV1_2,
+							MaxProtocolVersion: networking.ServerTLSSettings_TLSV1_3,
 						},
 					},
 				},
 			},
 		},
 		{
+			name: "complete config",
 			input: &networking.Gateway{
 				Servers: []*networking.Server{
 					{
@@ -144,24 +213,28 @@ func TestApplyGateway(t *testing.T) {
 					},
 					Mode:         networking.ServerTLSSettings_MUTUAL,
 					CipherSuites: []string{"ECDHE-RSA-AES256-GCM-SHA384"},
+					MinVersion:   "TLSv1.2",
+					MaxVersion:   "TLSv1.3",
 				},
 			},
 			expect: &networking.Gateway{
 				Servers: []*networking.Server{
-					{
-						Port: &networking.Port{
-							Protocol: "HTTPS",
-						},
+					{Port: &networking.Port{
+						Protocol: "HTTPS",
+					},
 						Tls: &networking.ServerTLSSettings{
-							CredentialName: "kubernetes-ingress://cluster/foo/bar",
-							Mode:           networking.ServerTLSSettings_MUTUAL,
-							CipherSuites:   []string{"ECDHE-RSA-AES256-GCM-SHA384"},
+							CredentialName:     "kubernetes-ingress://cluster/foo/bar",
+							Mode:               networking.ServerTLSSettings_MUTUAL,
+							CipherSuites:       []string{"ECDHE-RSA-AES256-GCM-SHA384"},
+							MinProtocolVersion: networking.ServerTLSSettings_TLSV1_2,
+							MaxProtocolVersion: networking.ServerTLSSettings_TLSV1_3,
 						},
 					},
 				},
 			},
 		},
 		{
+			name: "invalid TLS version",
 			input: &networking.Gateway{
 				Servers: []*networking.Server{
 					{
@@ -169,20 +242,15 @@ func TestApplyGateway(t *testing.T) {
 							Protocol: "HTTPS",
 						},
 						Tls: &networking.ServerTLSSettings{
-							Mode:           networking.ServerTLSSettings_SIMPLE,
-							CredentialName: "kubernetes-ingress://cluster/foo/bar",
+							Mode: networking.ServerTLSSettings_SIMPLE,
 						},
 					},
 				},
 			},
 			config: &Ingress{
 				DownstreamTLS: &DownstreamTLSConfig{
-					CASecretName: types.NamespacedName{
-						Namespace: "foo",
-						Name:      "bar-cacert",
-					},
-					Mode:         networking.ServerTLSSettings_MUTUAL,
-					CipherSuites: []string{"ECDHE-RSA-AES256-GCM-SHA384"},
+					MinVersion: "invalid",
+					MaxVersion: "invalid",
 				},
 			},
 			expect: &networking.Gateway{
@@ -192,48 +260,10 @@ func TestApplyGateway(t *testing.T) {
 							Protocol: "HTTPS",
 						},
 						Tls: &networking.ServerTLSSettings{
-							CredentialName: "kubernetes-ingress://cluster/foo/bar",
-							Mode:           networking.ServerTLSSettings_MUTUAL,
-							CipherSuites:   []string{"ECDHE-RSA-AES256-GCM-SHA384"},
-						},
-					},
-				},
-			},
-		},
-		{
-			input: &networking.Gateway{
-				Servers: []*networking.Server{
-					{
-						Port: &networking.Port{
-							Protocol: "HTTPS",
-						},
-						Tls: &networking.ServerTLSSettings{
-							Mode:           networking.ServerTLSSettings_SIMPLE,
-							CredentialName: "kubernetes-ingress://cluster/foo/bar",
-						},
-					},
-				},
-			},
-			config: &Ingress{
-				DownstreamTLS: &DownstreamTLSConfig{
-					CASecretName: types.NamespacedName{
-						Namespace: "bar",
-						Name:      "foo",
-					},
-					Mode:         networking.ServerTLSSettings_MUTUAL,
-					CipherSuites: []string{"ECDHE-RSA-AES256-GCM-SHA384"},
-				},
-			},
-			expect: &networking.Gateway{
-				Servers: []*networking.Server{
-					{
-						Port: &networking.Port{
-							Protocol: "HTTPS",
-						},
-						Tls: &networking.ServerTLSSettings{
-							CredentialName: "kubernetes-ingress://cluster/foo/bar",
-							Mode:           networking.ServerTLSSettings_SIMPLE,
-							CipherSuites:   []string{"ECDHE-RSA-AES256-GCM-SHA384"},
+							Mode: networking.ServerTLSSettings_SIMPLE,
+							// Invalid versions should default to TLS_AUTO
+							MinProtocolVersion: networking.ServerTLSSettings_TLS_AUTO,
+							MaxProtocolVersion: networking.ServerTLSSettings_TLS_AUTO,
 						},
 					},
 				},
@@ -241,11 +271,59 @@ func TestApplyGateway(t *testing.T) {
 		},
 	}

-	for _, testCase := range testCases {
-		t.Run("", func(t *testing.T) {
-			parser.ApplyGateway(testCase.input, testCase.config)
-			if !reflect.DeepEqual(testCase.input, testCase.expect) {
-				t.Fatalf("Should be equal")
+	for _, tc := range testCases {
+		t.Run(tc.name, func(t *testing.T) {
+			parser.ApplyGateway(tc.input, tc.config)
+			if !reflect.DeepEqual(tc.input, tc.expect) {
+				t.Fatalf("ApplyGateway result mismatch for %s:\nExpect: %+v\nGot: %+v",
+					tc.name, tc.expect, tc.input)
+			}
+		})
+	}
+}
+
+func TestNeedDownstreamTLS(t *testing.T) {
+	testCases := []struct {
+		name        string
+		annotations map[string]string
+		expect      bool
+	}{
+		{
+			name:        "empty annotations",
+			annotations: map[string]string{},
+			expect:      false,
+		},
+		{
+			name: "with ssl cipher",
+			annotations: map[string]string{
+				buildNginxAnnotationKey(sslCipher): "ECDHE-RSA-AES256-GCM-SHA384",
+			},
+			expect: true,
+		},
+		{
+			name: "with TLS version",
+			annotations: map[string]string{
+				buildNginxAnnotationKey(annotationMinTLSVersion): "TLSv1.2",
+			},
+			expect: true,
+		},
+		{
+			name: "with multiple TLS configs",
+			annotations: map[string]string{
+				buildNginxAnnotationKey(sslCipher):               "ECDHE-RSA-AES256-GCM-SHA384",
+				buildNginxAnnotationKey(annotationMinTLSVersion): "TLSv1.2",
+				buildNginxAnnotationKey(annotationMaxTLSVersion): "TLSv1.3",
+			},
+			expect: true,
+		},
+	}
+
+	for _, tc := range testCases {
+		t.Run(tc.name, func(t *testing.T) {
+			result := needDownstreamTLS(tc.annotations)
+			if result != tc.expect {
+				t.Errorf("needDownstreamTLS() for %s = %v, want %v",
+					tc.name, result, tc.expect)
 			}
 		})
 	}
--- a/pkg/ingress/kube/ingress/status.go
+++ b/pkg/ingress/kube/ingress/status.go
@@ -81,8 +81,6 @@ func (s *statusSyncer) runUpdateStatus() error {
 		return err
 	}

-	IngressLog.Debugf("found number %d of svc", len(svcList))
-
 	lbStatusList := common.GetLbStatusListV1Beta1(svcList)
 	if len(lbStatusList) == 0 {
 		return nil
--- a/pkg/ingress/kube/ingressv1/controller.go
+++ b/pkg/ingress/kube/ingressv1/controller.go
@@ -162,6 +162,7 @@ func (c *controller) onEvent(namespacedName types.NamespacedName) error {
 			delete(c.ingresses, namespacedName.String())
 			c.mutex.Unlock()
 		} else {
+			IngressLog.Warnf("ingressLister Get failed, ingress: %s, err: %v", namespacedName, err)
 			return err
 		}
 	}
@@ -171,7 +172,7 @@ func (c *controller) onEvent(namespacedName types.NamespacedName) error {
 		return nil
 	}

-	IngressLog.Debugf("ingress: %s, event: %s", namespacedName, event)
+	IngressLog.Infof("ingress: %s, event: %s", namespacedName, event)

 	// we should check need process only when event is not delete,
 	// if it is delete event, and previously processed, we need to process too.
@@ -181,7 +182,7 @@ func (c *controller) onEvent(namespacedName types.NamespacedName) error {
 			return err
 		}
 		if !shouldProcess {
-			IngressLog.Infof("no need process, ingress %s", namespacedName)
+			IngressLog.Infof("no need process, ingress: %s", namespacedName)
 			return nil
 		}
 	}
@@ -279,10 +280,17 @@ func (c *controller) List() []config.Config {
 	for _, raw := range c.ingressInformer.Informer.GetStore().List() {
 		ing, ok := raw.(*ingress.Ingress)
 		if !ok {
+			IngressLog.Warnf("get ingress from informer failed: %v", raw)
 			continue
 		}

-		if should, err := c.shouldProcessIngress(ing); !should || err != nil {
+		should, err := c.shouldProcessIngress(ing)
+		if err != nil {
+			IngressLog.Warnf("check should process ingress failed: %v", err)
+			continue
+		}
+		if !should {
+			IngressLog.Debugf("no need process ingress: %s/%s", ing.Namespace, ing.Name)
 			continue
 		}

--- a/pkg/ingress/kube/ingressv1/status.go
+++ b/pkg/ingress/kube/ingressv1/status.go
@@ -81,8 +81,6 @@ func (s *statusSyncer) runUpdateStatus() error {
 		return err
 	}

-	IngressLog.Debugf("found number %d of svc", len(svcList))
-
 	lbStatusList := common.GetLbStatusListV1(svcList)
 	if len(lbStatusList) == 0 {
 		return nil
--- a/pkg/ingress/kube/kingress/status.go
+++ b/pkg/ingress/kube/kingress/status.go
@@ -77,7 +77,6 @@ func (s *statusSyncer) runUpdateStatus() error {
 		return err
 	}

-	IngressLog.Debugf("found number %d of svc", len(svcList))
 	lbStatusList := common2.GetLbStatusList(svcList)
 	return s.updateStatus(lbStatusList)
 }
--- a/pkg/ingress/log/log.go
+++ b/pkg/ingress/log/log.go
@@ -14,6 +14,6 @@

 package log

-import "istio.io/pkg/log"
+import "istio.io/istio/pkg/log"

-var IngressLog = log.RegisterScope("ingress", "Higress Ingress process.", 0)
+var IngressLog = log.RegisterScope("ingress", "Higress Ingress process.")
--- a/plugins/wasm-cpp/WORKSPACE
+++ b/plugins/wasm-cpp/WORKSPACE
@@ -27,13 +27,17 @@ http_archive(
    url = "https://github.com/higress-group/proxy-wasm-cpp-sdk/archive/" + PROXY_WASM_CPP_SDK_SHA + ".tar.gz",
 )

-load("@proxy_wasm_cpp_sdk//bazel/dep:deps.bzl", "wasm_dependencies")
+load("@proxy_wasm_cpp_sdk//bazel:repositories.bzl", "proxy_wasm_cpp_sdk_repositories")

-wasm_dependencies()
+proxy_wasm_cpp_sdk_repositories()

-load("@proxy_wasm_cpp_sdk//bazel/dep:deps_extra.bzl", "wasm_dependencies_extra")
+load("@proxy_wasm_cpp_sdk//bazel:dependencies.bzl", "proxy_wasm_cpp_sdk_dependencies")

-wasm_dependencies_extra()
+proxy_wasm_cpp_sdk_dependencies()
+
+load("@proxy_wasm_cpp_sdk//bazel:dependencies_extra.bzl", "proxy_wasm_cpp_sdk_dependencies_extra")
+
+proxy_wasm_cpp_sdk_dependencies_extra()

 load("@istio_ecosystem_wasm_extensions//bazel:wasm.bzl", "wasm_libraries")

--- a/plugins/wasm-cpp/bazel/absl.patch
+++ b/plugins/wasm-cpp/bazel/absl.patch
@@ -2,16 +2,16 @@ diff --git a/absl/time/internal/cctz/src/time_zone_format.cc b/absl/time/interna
 index d8cb047..0c5f182 100644
 --- a/absl/time/internal/cctz/src/time_zone_format.cc
 +++ b/absl/time/internal/cctz/src/time_zone_format.cc
-@@ -18,6 +18,8 @@
- #endif
- #endif
+@@ -12,6 +12,8 @@
+ //   See the License for the specific language governing permissions and
+ //   limitations under the License.
 
 +#define HAS_STRPTIME 0
 +
- #if defined(HAS_STRPTIME) && HAS_STRPTIME
- #if !defined(_XOPEN_SOURCE)
- #define _XOPEN_SOURCE  // Definedness suffices for strptime.
-@@ -58,7 +60,7 @@ namespace {
+ #if !defined(HAS_STRPTIME)
+ #if !defined(_MSC_VER) && !defined(__MINGW32__)
+ #define HAS_STRPTIME 1  // assume everyone has strptime() except windows
+@@ -58,7 +60,7 @@
 
 #if !HAS_STRPTIME
 // Build a strptime() using C++11's std::get_time().
@@ -20,7 +20,7 @@ index d8cb047..0c5f182 100644
   std::istringstream input(s);
   input >> std::get_time(tm, fmt);
   if (input.fail()) return nullptr;
-@@ -648,7 +650,7 @@ const char* ParseSubSeconds(const char* dp, detail::femtoseconds* subseconds) {
+@@ -648,7 +650,7 @@
 // Parses a string into a std::tm using strptime(3).
 const char* ParseTM(const char* dp, const char* fmt, std::tm* tm) {
   if (dp != nullptr) {
--- a/plugins/wasm-cpp/bazel/wasm.bzl
+++ b/plugins/wasm-cpp/bazel/wasm.bzl
@@ -9,9 +9,9 @@ load(
 def wasm_libraries():
    http_archive(
        name = "com_google_absl",
-        sha256 = "ec8ef47335310cc3382bdc0d0cc1097a001e67dc83fcba807845aa5696e7e1e4",
-        strip_prefix = "abseil-cpp-302b250e1d917ede77b5ff00a6fd9f28430f1563",
-        url = "https://github.com/abseil/abseil-cpp/archive/302b250e1d917ede77b5ff00a6fd9f28430f1563.tar.gz",
+        sha256 = "3a0bb3d2e6f53352526a8d1a7e7b5749c68cd07f2401766a404fb00d2853fa49",
+        strip_prefix = "abseil-cpp-4bbdb026899fea9f882a95cbd7d6a4adaf49b2dd",
+        url = "https://github.com/abseil/abseil-cpp/archive/4bbdb026899fea9f882a95cbd7d6a4adaf49b2dd.tar.gz",
        patch_args = ["-p1"],
        patches = ["//bazel:absl.patch"],
    )
@@ -33,8 +33,8 @@ def wasm_libraries():
        urls = ["https://github.com/google/googletest/archive/release-1.10.0.tar.gz"],
    )

-    PROXY_WASM_CPP_HOST_SHA = "7850d1721fe3dd2ccfb86a06116f76c23b1f1bf8"
-    PROXY_WASM_CPP_HOST_SHA256 = "740690fc1d749849f6e24b5bc48a07dabc0565a7d03b6cd13425dba693956c57"
+    PROXY_WASM_CPP_HOST_SHA = "ecf42a27fcf78f42e64037d4eff1a0ca5a61e403"
+    PROXY_WASM_CPP_HOST_SHA256 = "9748156731e9521837686923321bf12725c32c9fa8355218209831cc3ee87080"

    http_archive(
        name = "proxy_wasm_cpp_host",
--- a/plugins/wasm-cpp/common/http_util.cc
+++ b/plugins/wasm-cpp/common/http_util.cc
@@ -19,7 +19,6 @@
 #include "absl/strings/str_cat.h"
 #include "absl/strings/str_format.h"
 #include "absl/strings/str_split.h"
-
 #include "common/common_util.h"

 namespace Wasm::Common::Http {
@@ -190,7 +189,8 @@ std::vector<std::string> getAllOfHeader(std::string_view key) {
  std::vector<std::string> result;
  auto headers = getRequestHeaderPairs()->pairs();
  for (auto& header : headers) {
-    if (absl::EqualsIgnoreCase(Wasm::Common::stdToAbsl(header.first), Wasm::Common::stdToAbsl(key))) {
+    if (absl::EqualsIgnoreCase(Wasm::Common::stdToAbsl(header.first),
+                               Wasm::Common::stdToAbsl(key))) {
      result.push_back(std::string(header.second));
    }
  }
@@ -225,7 +225,8 @@ void forEachCookie(
        v = v.substr(1, v.size() - 2);
      }

-      if (!cookie_consumer(Wasm::Common::abslToStd(k), Wasm::Common::abslToStd(v))) {
+      if (!cookie_consumer(Wasm::Common::abslToStd(k),
+                           Wasm::Common::abslToStd(v))) {
        return;
      }
    }
@@ -265,7 +266,63 @@ std::string buildOriginalUri(std::optional<uint32_t> max_path_length) {
  auto scheme = scheme_ptr->view();
  auto host_ptr = getRequestHeader(Header::Host);
  auto host = host_ptr->view();
-  return absl::StrCat(Wasm::Common::stdToAbsl(scheme), "://", Wasm::Common::stdToAbsl(host),  Wasm::Common::stdToAbsl(final_path));
+  return absl::StrCat(Wasm::Common::stdToAbsl(scheme), "://",
+                      Wasm::Common::stdToAbsl(host),
+                      Wasm::Common::stdToAbsl(final_path));
+}
+
+void extractHostPathFromUri(const absl::string_view& uri,
+                            absl::string_view& host, absl::string_view& path) {
+  /**
+   *  URI RFC: https://www.ietf.org/rfc/rfc2396.txt
+   *
+   *  Example:
+   *  uri  = "https://example.com:8443/certs"
+   *  pos:         ^
+   *  host_pos:       ^
+   *  path_pos:                       ^
+   *  host = "example.com:8443"
+   *  path = "/certs"
+   */
+  const auto pos = uri.find("://");
+  // Start position of the host
+  const auto host_pos = (pos == std::string::npos) ? 0 : pos + 3;
+  // Start position of the path
+  const auto path_pos = uri.find('/', host_pos);
+  if (path_pos == std::string::npos) {
+    // If uri doesn't have "/", the whole string is treated as host.
+    host = uri.substr(host_pos);
+    path = "/";
+  } else {
+    host = uri.substr(host_pos, path_pos - host_pos);
+    path = uri.substr(path_pos);
+  }
+}
+
+void extractPathWithoutArgsFromUri(const std::string_view& uri,
+                                   std::string_view& path_without_args) {
+  auto params_pos = uri.find('?');
+  size_t uri_end;
+  if (params_pos == std::string::npos) {
+    uri_end = uri.size();
+  } else {
+    uri_end = params_pos;
+  }
+  path_without_args = uri.substr(0, uri_end);
+}
+
+bool hasRequestBody() {
+  auto contentType = getRequestHeader("content-type")->toString();
+  auto contentLengthStr = getRequestHeader("content-length")->toString();
+  auto transferEncoding = getRequestHeader("transfer-encoding")->toString();
+
+  if (!contentType.empty()) {
+    return true;
+  }
+  if (!contentLengthStr.empty()) {
+    return true;
+  }
+  return transferEncoding.find("chunked") != std::string::npos;
 }

 }  // namespace Wasm::Common::Http
--- a/plugins/wasm-cpp/common/http_util.h
+++ b/plugins/wasm-cpp/common/http_util.h
@@ -42,6 +42,12 @@ namespace Wasm::Common::Http {
 using QueryParams = std::map<std::string, std::string>;
 using SystemTime = std::chrono::time_point<std::chrono::system_clock>;

+namespace Status {
+constexpr int OK = 200;
+constexpr int InternalServerError = 500;
+constexpr int Unauthorized = 401;
+}  // namespace Status
+
 namespace Header {
 constexpr std::string_view Scheme(":scheme");
 constexpr std::string_view Method(":method");
@@ -52,14 +58,17 @@ constexpr std::string_view Accept("accept");
 constexpr std::string_view ContentMD5("content-md5");
 constexpr std::string_view ContentType("content-type");
 constexpr std::string_view ContentLength("content-length");
+constexpr std::string_view TransferEncoding("transfer-encoding");
 constexpr std::string_view UserAgent("user-agent");
 constexpr std::string_view Date("date");
 constexpr std::string_view Cookie("cookie");
+constexpr std::string_view StrictTransportSecurity("strict-transport-security");
 }  // namespace Header

 namespace ContentTypeValues {
 constexpr std::string_view Grpc{"application/grpc"};
-}
+constexpr std::string_view Json{"application/json"};
+}  // namespace ContentTypeValues

 class PercentEncoding {
 public:
@@ -142,4 +151,10 @@ std::unordered_map<std::string, std::string> parseCookies(

 std::string buildOriginalUri(std::optional<uint32_t> max_path_length);

+void extractHostPathFromUri(const absl::string_view& uri,
+                            absl::string_view& host, absl::string_view& path);
+
+void extractPathWithoutArgsFromUri(const std::string_view& uri,
+                                   std::string_view& path_without_args);
+bool hasRequestBody();
 }  // namespace Wasm::Common::Http
--- a/plugins/wasm-cpp/extensions/bot_detect/BUILD
+++ b/plugins/wasm-cpp/extensions/bot_detect/BUILD
@@ -12,10 +12,10 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.

-load("@proxy_wasm_cpp_sdk//bazel/wasm:wasm.bzl", "wasm_cc_binary")
+load("@proxy_wasm_cpp_sdk//bazel:defs.bzl", "proxy_wasm_cc_binary")
 load("//bazel:wasm.bzl", "declare_wasm_image_targets")

-wasm_cc_binary(
+proxy_wasm_cc_binary(
    name = "bot_detect.wasm",
    srcs = [
        "plugin.cc",
@@ -28,7 +28,6 @@ wasm_cc_binary(
        "//common:http_util",
        "//common:regex_util",
        "//common:rule_util",
-        "@proxy_wasm_cpp_sdk//:proxy_wasm_intrinsics",
    ],
 )

--- a/plugins/wasm-cpp/extensions/custom_response/BUILD
+++ b/plugins/wasm-cpp/extensions/custom_response/BUILD
@@ -12,10 +12,10 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.

-load("@proxy_wasm_cpp_sdk//bazel/wasm:wasm.bzl", "wasm_cc_binary")
+load("@proxy_wasm_cpp_sdk//bazel:defs.bzl", "proxy_wasm_cc_binary")
 load("//bazel:wasm.bzl", "declare_wasm_image_targets")

-wasm_cc_binary(
+proxy_wasm_cc_binary(
    name = "custom_response.wasm",
    srcs = [
        "plugin.cc",
@@ -27,7 +27,6 @@ wasm_cc_binary(
        "//common:json_util",
        "//common:http_util",
        "//common:rule_util",
-        "@proxy_wasm_cpp_sdk//:proxy_wasm_intrinsics",
    ],
 )

--- a/plugins/wasm-cpp/extensions/hmac_auth/BUILD
+++ b/plugins/wasm-cpp/extensions/hmac_auth/BUILD
@@ -12,10 +12,10 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.

-load("@proxy_wasm_cpp_sdk//bazel/wasm:wasm.bzl", "wasm_cc_binary")
+load("@proxy_wasm_cpp_sdk//bazel:defs.bzl", "proxy_wasm_cc_binary")
 load("//bazel:wasm.bzl", "declare_wasm_image_targets")

-wasm_cc_binary(
+proxy_wasm_cc_binary(
    name = "hmac_auth.wasm",
    srcs = [
        "plugin.cc",
@@ -30,7 +30,6 @@ wasm_cc_binary(
        "//common:crypto_util",
        "//common:http_util",
        "//common:rule_util",
-        "@proxy_wasm_cpp_sdk//:proxy_wasm_intrinsics",
    ],
 )

--- a/plugins/wasm-cpp/extensions/jwt_auth/BUILD
+++ b/plugins/wasm-cpp/extensions/jwt_auth/BUILD
@@ -12,7 +12,7 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.

-load("@proxy_wasm_cpp_sdk//bazel/wasm:wasm.bzl", "wasm_cc_binary")
+load("@proxy_wasm_cpp_sdk//bazel:defs.bzl", "proxy_wasm_cc_binary")
 load("//bazel:wasm.bzl", "declare_wasm_image_targets")

 wasm_cc_binary(
@@ -33,7 +33,6 @@ wasm_cc_binary(
        "//common:json_util",
        "//common:http_util",
        "//common:rule_util",
-        "@proxy_wasm_cpp_sdk//:proxy_wasm_intrinsics",
    ],
 )

--- a/plugins/wasm-cpp/extensions/key_auth/BUILD
+++ b/plugins/wasm-cpp/extensions/key_auth/BUILD
@@ -12,10 +12,10 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.

-load("@proxy_wasm_cpp_sdk//bazel/wasm:wasm.bzl", "wasm_cc_binary")
+load("@proxy_wasm_cpp_sdk//bazel:defs.bzl", "proxy_wasm_cc_binary")
 load("//bazel:wasm.bzl", "declare_wasm_image_targets")

-wasm_cc_binary(
+proxy_wasm_cc_binary(
    name = "key_auth.wasm",
    srcs = [
        "plugin.cc",
@@ -28,7 +28,6 @@ wasm_cc_binary(
        "//common:json_util",
        "//common:http_util",
        "//common:rule_util",
-        "@proxy_wasm_cpp_sdk//:proxy_wasm_intrinsics",
    ],
 )

--- a/plugins/wasm-cpp/extensions/key_rate_limit/BUILD
+++ b/plugins/wasm-cpp/extensions/key_rate_limit/BUILD
@@ -12,10 +12,10 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.

-load("@proxy_wasm_cpp_sdk//bazel/wasm:wasm.bzl", "wasm_cc_binary")
+load("@proxy_wasm_cpp_sdk//bazel:defs.bzl", "proxy_wasm_cc_binary")
 load("//bazel:wasm.bzl", "declare_wasm_image_targets")

-wasm_cc_binary(
+proxy_wasm_cc_binary(
    name = "key_rate_limit.wasm",
    srcs = [
        "plugin.cc",
@@ -29,7 +29,6 @@ wasm_cc_binary(
        "//common:json_util",
        "//common:http_util",
        "//common:rule_util",
-        "@proxy_wasm_cpp_sdk//:proxy_wasm_intrinsics",
    ],
 )

--- a/plugins/wasm-cpp/extensions/model_mapper/BUILD
+++ b/plugins/wasm-cpp/extensions/model_mapper/BUILD
@@ -0,0 +1,70 @@
+# Copyright (c) 2022 Alibaba Group Holding Ltd.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+load("@proxy_wasm_cpp_sdk//bazel:defs.bzl", "proxy_wasm_cc_binary")
+load("//bazel:wasm.bzl", "declare_wasm_image_targets")
+
+proxy_wasm_cc_binary(
+    name = "model_mapper.wasm",
+    srcs = [
+        "plugin.cc",
+        "plugin.h",
+    ],
+    deps = [
+        "@proxy_wasm_cpp_sdk//:proxy_wasm_intrinsics_higress",
+        "@com_google_absl//absl/strings",
+        "@com_google_absl//absl/time",
+        "//common:json_util",
+        "//common:http_util",
+        "//common:rule_util",
+    ],
+)
+
+cc_library(
+    name = "model_mapper_lib",
+    srcs = [
+        "plugin.cc",
+    ],
+    hdrs = [
+        "plugin.h",
+    ],
+    copts = ["-DNULL_PLUGIN"],
+    deps = [
+        "@com_google_absl//absl/strings",
+        "@com_google_absl//absl/time",
+        "//common:json_util",
+        "@proxy_wasm_cpp_host//:lib",
+        "//common:http_util_nullvm",
+        "//common:rule_util_nullvm",
+    ],
+)
+
+cc_test(
+    name = "model_mapper_test",
+    srcs = [
+        "plugin_test.cc",
+    ],
+    copts = ["-DNULL_PLUGIN"],
+    deps = [
+        ":model_mapper_lib",
+        "@com_google_googletest//:gtest",
+        "@com_google_googletest//:gtest_main",
+        "@proxy_wasm_cpp_host//:lib",
+    ],
+)
+
+declare_wasm_image_targets(
+    name = "model_mapper",
+    wasm_file = ":model_mapper.wasm",
+)
--- a/plugins/wasm-cpp/extensions/model_mapper/README.md
+++ b/plugins/wasm-cpp/extensions/model_mapper/README.md
@@ -0,0 +1,63 @@
+## 功能说明
+`model-mapper`插件实现了基于LLM协议中的model参数路由的功能
+
+## 配置字段
+
+| 名称                 | 数据类型        | 填写要求                | 默认值                   | 描述                                                                                                                                                                                                                                                         |
+| -----------          | --------------- | ----------------------- | ------                   | -------------------------------------------                                                                                                                                                                                                                  |
+| `modelKey`           | string          | 选填                    | model                    | 请求body中model参数的位置                                                                                                                                                                                                                                    |
+| `modelMapping`       | map of string   | 选填                    | -                        | AI 模型映射表，用于将请求中的模型名称映射为服务提供商支持模型名称。<br/>1. 支持前缀匹配。例如用 "gpt-3-*" 匹配所有名称以“gpt-3-”开头的模型；<br/>2. 支持使用 "*" 为键来配置通用兜底映射关系；<br/>3. 如果映射的目标名称为空字符串 ""，则表示保留原模型名称。 |
+| `enableOnPathSuffix` | array of string | 选填                    | ["/v1/chat/completions"] | 只对这些特定路径后缀的请求生效                                                                                                                              ## 运行属性
+
+插件执行阶段：认证阶段
+插件执行优先级：800
+                                                                                                 |
+## 效果说明
+
+如下配置
+
+```yaml
+modelMapping:
+  'gpt-4-*': "qwen-max"
+  'gpt-4o': "qwen-vl-plus"
+  '*': "qwen-turbo"
+```
+
+开启后，`gpt-4-` 开头的模型参数会被改写为 `qwen-max`, `gpt-4o` 会被改写为 `qwen-vl-plus`，其他所有模型会被改写为 `qwen-turbo`
+
+例如原本的请求是：
+
+```json
+{
+    "model": "gpt-4o",
+    "frequency_penalty": 0,
+    "max_tokens": 800,
+    "stream": false,
+    "messages": [{
+        "role": "user",
+        "content": "higress项目主仓库的github地址是什么"
+    }],
+    "presence_penalty": 0,
+    "temperature": 0.7,
+    "top_p": 0.95
+}
+```
+
+
+经过这个插件后，原始的 LLM 请求体将被改成：
+
+```json
+{
+    "model": "qwen-vl-plus",
+    "frequency_penalty": 0,
+    "max_tokens": 800,
+    "stream": false,
+    "messages": [{
+        "role": "user",
+        "content": "higress项目主仓库的github地址是什么"
+    }],
+    "presence_penalty": 0,
+    "temperature": 0.7,
+    "top_p": 0.95
+}
+```
--- a/plugins/wasm-cpp/extensions/model_mapper/README_EN.md
+++ b/plugins/wasm-cpp/extensions/model_mapper/README_EN.md
@@ -0,0 +1,65 @@
+## Function Description
+The `model-mapper` plugin implements the functionality of routing based on the model parameter in the LLM protocol.
+
+## Configuration Fields
+
+| Name                 | Data Type        | Filling Requirement                | Default Value                   | Description                                                                                                                                                                                                                                                         |
+| -----------          | --------------- | -----------------------            | ------                          | -------------------------------------------                                                                                                                                                                                                                  |
+| `modelKey`           | string          | Optional                           | model                           | The location of the model parameter in the request body.                                                                                                                                                                                                            |
+| `modelMapping`       | map of string   | Optional                           | -                               | AI model mapping table, used to map the model names in the request to the model names supported by the service provider.<br/>1. Supports prefix matching. For example, use "gpt-3-*" to match all models whose names start with “gpt-3-”;<br/>2. Supports using "*" as the key to configure a generic fallback mapping relationship;<br/>3. If the target name in the mapping is an empty string "", it means to keep the original model name. |
+| `enableOnPathSuffix` | array of string | Optional                           | ["/v1/chat/completions"]        | Only applies to requests with these specific path suffixes.                                                                                                                                           |
+
+## Runtime Properties
+
+Plugin execution phase: Authentication phase
+Plugin execution priority: 800
+
+## Effect Description
+
+With the following configuration:
+
+```yaml
+modelMapping:
+  'gpt-4-*': "qwen-max"
+  'gpt-4o': "qwen-vl-plus"
+  '*': "qwen-turbo"
+```
+
+After enabling, model parameters starting with `gpt-4-` will be rewritten to `qwen-max`, `gpt-4o` will be rewritten to `qwen-vl-plus`, and all other models will be rewritten to `qwen-turbo`.
+
+For example, if the original request was:
+
+```json
+{
+    "model": "gpt-4o",
+    "frequency_penalty": 0,
+    "max_tokens": 800,
+    "stream": false,
+    "messages": [{
+        "role": "user",
+        "content": "What is the GitHub address of the main repository for the higress project?"
+    }],
+    "presence_penalty": 0,
+    "temperature": 0.7,
+    "top_p": 0.95
+}
+```
+
+
+After processing by this plugin, the original LLM request body will be modified to:
+
+```json
+{
+    "model": "qwen-vl-plus",
+    "frequency_penalty": 0,
+    "max_tokens": 800,
+    "stream": false,
+    "messages": [{
+        "role": "user",
+        "content": "What is the GitHub address of the main repository for the higress project?"
+    }],
+    "presence_penalty": 0,
+    "temperature": 0.7,
+    "top_p": 0.95
+}
+```
--- a/plugins/wasm-cpp/extensions/model_mapper/plugin.cc
+++ b/plugins/wasm-cpp/extensions/model_mapper/plugin.cc
@@ -0,0 +1,243 @@
+// Copyright (c) 2022 Alibaba Group Holding Ltd.
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+//      http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+#include "extensions/model_mapper/plugin.h"
+
+#include <array>
+#include <limits>
+
+#include "absl/strings/str_cat.h"
+#include "absl/strings/str_split.h"
+#include "common/http_util.h"
+#include "common/json_util.h"
+
+using ::nlohmann::json;
+using ::Wasm::Common::JsonArrayIterate;
+using ::Wasm::Common::JsonGetField;
+using ::Wasm::Common::JsonObjectIterate;
+using ::Wasm::Common::JsonValueAs;
+
+#ifdef NULL_PLUGIN
+
+namespace proxy_wasm {
+namespace null_plugin {
+namespace model_mapper {
+
+PROXY_WASM_NULL_PLUGIN_REGISTRY
+
+#endif
+
+static RegisterContextFactory register_ModelMapper(
+    CONTEXT_FACTORY(PluginContext), ROOT_FACTORY(PluginRootContext));
+
+namespace {
+
+constexpr std::string_view SetDecoderBufferLimitKey =
+    "SetRequestBodyBufferLimit";
+constexpr std::string_view DefaultMaxBodyBytes = "10485760";
+
+}  // namespace
+
+bool PluginRootContext::parsePluginConfig(const json& configuration,
+                                          ModelMapperConfigRule& rule) {
+  if (auto it = configuration.find("modelKey"); it != configuration.end()) {
+    if (it->is_string()) {
+      rule.model_key_ = it->get<std::string>();
+    } else {
+      LOG_ERROR("Invalid type for modelKey. Expected string.");
+      return false;
+    }
+  }
+
+  if (auto it = configuration.find("modelMapping"); it != configuration.end()) {
+    if (!it->is_object()) {
+      LOG_ERROR("Invalid type for modelMapping. Expected object.");
+      return false;
+    }
+    auto model_mapping = it->get<Wasm::Common::JsonObject>();
+    if (!JsonObjectIterate(model_mapping, [&](std::string key) -> bool {
+          auto model_json = model_mapping.find(key);
+          if (!model_json->is_string()) {
+            LOG_ERROR(
+                "Invalid type for item in modelMapping. Expected string.");
+            return false;
+          }
+          if (key == "*") {
+            rule.default_model_mapping_ = model_json->get<std::string>();
+            return true;
+          }
+          if (absl::EndsWith(key, "*")) {
+            rule.prefix_model_mapping_.emplace_back(
+                absl::StripSuffix(key, "*"), model_json->get<std::string>());
+            return true;
+          }
+          auto ret = rule.exact_model_mapping_.emplace(
+              key, model_json->get<std::string>());
+          if (!ret.second) {
+            LOG_ERROR("Duplicate key in modelMapping: " + key);
+            return false;
+          }
+          return true;
+        })) {
+      return false;
+    }
+  }
+
+  if (!JsonArrayIterate(
+          configuration, "enableOnPathSuffix", [&](const json& item) -> bool {
+            if (item.is_string()) {
+              rule.enable_on_path_suffix_.emplace_back(item.get<std::string>());
+              return true;
+            }
+            return false;
+          })) {
+    LOG_WARN("Invalid type for item in enableOnPathSuffix. Expected string.");
+    return false;
+  }
+  return true;
+}
+
+bool PluginRootContext::onConfigure(size_t size) {
+  // Parse configuration JSON string.
+  if (size > 0 && !configure(size)) {
+    LOG_WARN("configuration has errors initialization will not continue.");
+    return false;
+  }
+  return true;
+}
+
+bool PluginRootContext::configure(size_t configuration_size) {
+  auto configuration_data = getBufferBytes(WasmBufferType::PluginConfiguration,
+                                           0, configuration_size);
+  // Parse configuration JSON string.
+  auto result = ::Wasm::Common::JsonParse(configuration_data->view());
+  if (!result) {
+    LOG_WARN(absl::StrCat("cannot parse plugin configuration JSON string: ",
+                          configuration_data->view()));
+    return false;
+  }
+  if (!parseRuleConfig(result.value())) {
+    LOG_WARN(absl::StrCat("cannot parse plugin configuration JSON string: ",
+                          configuration_data->view()));
+    return false;
+  }
+  return true;
+}
+
+FilterHeadersStatus PluginRootContext::onHeader(
+    const ModelMapperConfigRule& rule) {
+  if (!Wasm::Common::Http::hasRequestBody()) {
+    return FilterHeadersStatus::Continue;
+  }
+  auto path = getRequestHeader(Wasm::Common::Http::Header::Path)->toString();
+  auto params_pos = path.find('?');
+  size_t uri_end;
+  if (params_pos == std::string::npos) {
+    uri_end = path.size();
+  } else {
+    uri_end = params_pos;
+  }
+  bool enable = false;
+  for (const auto& enable_suffix : rule.enable_on_path_suffix_) {
+    if (absl::EndsWith({path.c_str(), uri_end}, enable_suffix)) {
+      enable = true;
+      break;
+    }
+  }
+  if (!enable) {
+    return FilterHeadersStatus::Continue;
+  }
+  auto content_type_value =
+      getRequestHeader(Wasm::Common::Http::Header::ContentType);
+  if (!absl::StrContains(content_type_value->view(),
+                         Wasm::Common::Http::ContentTypeValues::Json)) {
+    return FilterHeadersStatus::Continue;
+  }
+  removeRequestHeader(Wasm::Common::Http::Header::ContentLength);
+  setFilterState(SetDecoderBufferLimitKey, DefaultMaxBodyBytes);
+  return FilterHeadersStatus::StopIteration;
+}
+
+FilterDataStatus PluginRootContext::onBody(const ModelMapperConfigRule& rule,
+                                           std::string_view body) {
+  const auto& exact_model_mapping = rule.exact_model_mapping_;
+  const auto& prefix_model_mapping = rule.prefix_model_mapping_;
+  const auto& default_model_mapping = rule.default_model_mapping_;
+  const auto& model_key = rule.model_key_;
+  auto body_json_opt = ::Wasm::Common::JsonParse(body);
+  if (!body_json_opt) {
+    LOG_WARN(absl::StrCat("cannot parse body to JSON string: ", body));
+    return FilterDataStatus::Continue;
+  }
+  auto body_json = body_json_opt.value();
+  std::string old_model;
+  if (body_json.contains(model_key)) {
+    old_model = body_json[model_key];
+  }
+  std::string model =
+      default_model_mapping.empty() ? old_model : default_model_mapping;
+  if (auto it = exact_model_mapping.find(old_model);
+      it != exact_model_mapping.end()) {
+    model = it->second;
+  } else {
+    for (auto& prefix_model_pair : prefix_model_mapping) {
+      if (absl::StartsWith(old_model, prefix_model_pair.first)) {
+        model = prefix_model_pair.second;
+        break;
+      }
+    }
+  }
+  if (!model.empty() && model != old_model) {
+    body_json[model_key] = model;
+    setBuffer(WasmBufferType::HttpRequestBody, 0,
+              std::numeric_limits<size_t>::max(), body_json.dump());
+    LOG_DEBUG(
+        absl::StrCat("model mapped, before:", old_model, ", after:", model));
+  }
+  return FilterDataStatus::Continue;
+}
+
+FilterHeadersStatus PluginContext::onRequestHeaders(uint32_t, bool) {
+  auto* rootCtx = rootContext();
+  return rootCtx->onHeaders([rootCtx, this](const auto& config) {
+    auto ret = rootCtx->onHeader(config);
+    if (ret == FilterHeadersStatus::StopIteration) {
+      this->config_ = &config;
+    }
+    return ret;
+  });
+}
+
+FilterDataStatus PluginContext::onRequestBody(size_t body_size,
+                                              bool end_stream) {
+  if (config_ == nullptr) {
+    return FilterDataStatus::Continue;
+  }
+  body_total_size_ += body_size;
+  if (!end_stream) {
+    return FilterDataStatus::StopIterationAndBuffer;
+  }
+  auto body =
+      getBufferBytes(WasmBufferType::HttpRequestBody, 0, body_total_size_);
+  auto* rootCtx = rootContext();
+  return rootCtx->onBody(*config_, body->view());
+}
+
+#ifdef NULL_PLUGIN
+
+}  // namespace model_mapper
+}  // namespace null_plugin
+}  // namespace proxy_wasm
+
+#endif
--- a/plugins/wasm-cpp/extensions/model_mapper/plugin.h
+++ b/plugins/wasm-cpp/extensions/model_mapper/plugin.h
@@ -0,0 +1,87 @@
+/*
+ * Copyright (c) 2022 Alibaba Group Holding Ltd.
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ *      http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+#include <assert.h>
+
+#include <string>
+#include <unordered_set>
+
+#include "common/route_rule_matcher.h"
+#define ASSERT(_X) assert(_X)
+
+#ifndef NULL_PLUGIN
+
+#include "proxy_wasm_intrinsics.h"
+
+#else
+
+#include "include/proxy-wasm/null_plugin.h"
+
+namespace proxy_wasm {
+namespace null_plugin {
+namespace model_mapper {
+
+#endif
+
+struct ModelMapperConfigRule {
+  std::string model_key_ = "model";
+  std::map<std::string, std::string> exact_model_mapping_;
+  std::vector<std::pair<std::string, std::string>> prefix_model_mapping_;
+  std::string default_model_mapping_;
+  std::vector<std::string> enable_on_path_suffix_ = {"/v1/chat/completions"};
+};
+
+// PluginRootContext is the root context for all streams processed by the
+// thread. It has the same lifetime as the worker thread and acts as target for
+// interactions that outlives individual stream, e.g. timer, async calls.
+class PluginRootContext : public RootContext,
+                          public RouteRuleMatcher<ModelMapperConfigRule> {
+ public:
+  PluginRootContext(uint32_t id, std::string_view root_id)
+      : RootContext(id, root_id) {}
+  ~PluginRootContext() {}
+  bool onConfigure(size_t) override;
+  FilterHeadersStatus onHeader(const ModelMapperConfigRule&);
+  FilterDataStatus onBody(const ModelMapperConfigRule&, std::string_view);
+  bool configure(size_t);
+
+ private:
+  bool parsePluginConfig(const json&, ModelMapperConfigRule&) override;
+};
+
+// Per-stream context.
+class PluginContext : public Context {
+ public:
+  explicit PluginContext(uint32_t id, RootContext* root) : Context(id, root) {}
+  FilterHeadersStatus onRequestHeaders(uint32_t, bool) override;
+  FilterDataStatus onRequestBody(size_t, bool) override;
+
+ private:
+  inline PluginRootContext* rootContext() {
+    return dynamic_cast<PluginRootContext*>(this->root());
+  }
+
+  size_t body_total_size_ = 0;
+  const ModelMapperConfigRule* config_ = nullptr;
+};
+
+#ifdef NULL_PLUGIN
+
+}  // namespace model_mapper
+}  // namespace null_plugin
+}  // namespace proxy_wasm
+
+#endif
--- a/plugins/wasm-cpp/extensions/model_mapper/plugin_test.cc
+++ b/plugins/wasm-cpp/extensions/model_mapper/plugin_test.cc
@@ -0,0 +1,301 @@
+// Copyright (c) 2022 Alibaba Group Holding Ltd.
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+//      http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+#include "extensions/model_mapper/plugin.h"
+
+#include <cstddef>
+
+#include "gmock/gmock.h"
+#include "gtest/gtest.h"
+#include "include/proxy-wasm/context.h"
+#include "include/proxy-wasm/null.h"
+
+namespace proxy_wasm {
+namespace null_plugin {
+namespace model_mapper {
+
+NullPluginRegistry* context_registry_;
+RegisterNullVmPluginFactory register_model_mapper_plugin("model_mapper", []() {
+  return std::make_unique<NullPlugin>(model_mapper::context_registry_);
+});
+
+class MockContext : public proxy_wasm::ContextBase {
+ public:
+  MockContext(WasmBase* wasm) : ContextBase(wasm) {}
+  MOCK_METHOD(BufferInterface*, getBuffer, (WasmBufferType));
+  MOCK_METHOD(WasmResult, log, (uint32_t, std::string_view));
+  MOCK_METHOD(WasmResult, setBuffer,
+              (WasmBufferType, size_t, size_t, std::string_view));
+  MOCK_METHOD(WasmResult, getHeaderMapValue,
+              (WasmHeaderMapType /* type */, std::string_view /* key */,
+               std::string_view* /*result */));
+  MOCK_METHOD(WasmResult, replaceHeaderMapValue,
+              (WasmHeaderMapType /* type */, std::string_view /* key */,
+               std::string_view /* value */));
+  MOCK_METHOD(WasmResult, removeHeaderMapValue,
+              (WasmHeaderMapType /* type */, std::string_view /* key */));
+  MOCK_METHOD(WasmResult, addHeaderMapValue,
+              (WasmHeaderMapType, std::string_view, std::string_view));
+  MOCK_METHOD(WasmResult, getProperty, (std::string_view, std::string*));
+  MOCK_METHOD(WasmResult, setProperty, (std::string_view, std::string_view));
+};
+class ModelMapperTest : public ::testing::Test {
+ protected:
+  ModelMapperTest() {
+    // Initialize test VM
+    test_vm_ = createNullVm();
+    wasm_base_ = std::make_unique<WasmBase>(
+        std::move(test_vm_), "test-vm", "", "",
+        std::unordered_map<std::string, std::string>{},
+        AllowedCapabilitiesMap{});
+    wasm_base_->load("model_mapper");
+    wasm_base_->initialize();
+    // Initialize host side context
+    mock_context_ = std::make_unique<MockContext>(wasm_base_.get());
+    current_context_ = mock_context_.get();
+    // Initialize Wasm sandbox context
+    root_context_ = std::make_unique<PluginRootContext>(0, "");
+    context_ = std::make_unique<PluginContext>(1, root_context_.get());
+
+    ON_CALL(*mock_context_, log(testing::_, testing::_))
+        .WillByDefault([](uint32_t, std::string_view m) {
+          std::cerr << m << "\n";
+          return WasmResult::Ok;
+        });
+
+    ON_CALL(*mock_context_, getBuffer(testing::_))
+        .WillByDefault([&](WasmBufferType type) {
+          if (type == WasmBufferType::HttpRequestBody) {
+            return &body_;
+          }
+          return &config_;
+        });
+    ON_CALL(*mock_context_, getHeaderMapValue(WasmHeaderMapType::RequestHeaders,
+                                              testing::_, testing::_))
+        .WillByDefault([&](WasmHeaderMapType, std::string_view header,
+                           std::string_view* result) {
+          if (header == "content-type") {
+            *result = "application/json";
+          } else if (header == "content-length") {
+            *result = "1024";
+          } else if (header == ":path") {
+            *result = path_;
+          }
+          return WasmResult::Ok;
+        });
+    ON_CALL(*mock_context_,
+            replaceHeaderMapValue(WasmHeaderMapType::RequestHeaders, testing::_,
+                                  testing::_))
+        .WillByDefault([&](WasmHeaderMapType, std::string_view key,
+                           std::string_view value) { return WasmResult::Ok; });
+    ON_CALL(*mock_context_,
+            removeHeaderMapValue(WasmHeaderMapType::RequestHeaders, testing::_))
+        .WillByDefault([&](WasmHeaderMapType, std::string_view key) {
+          return WasmResult::Ok;
+        });
+    ON_CALL(*mock_context_, addHeaderMapValue(WasmHeaderMapType::RequestHeaders,
+                                              testing::_, testing::_))
+        .WillByDefault([&](WasmHeaderMapType, std::string_view header,
+                           std::string_view value) { return WasmResult::Ok; });
+    ON_CALL(*mock_context_, getProperty(testing::_, testing::_))
+        .WillByDefault([&](std::string_view path, std::string* result) {
+          if (absl::StartsWith(path, "route_name")) {
+            *result = route_name_;
+          } else if (absl::StartsWith(path, "cluster_name")) {
+            *result = service_name_;
+          }
+          return WasmResult::Ok;
+        });
+    ON_CALL(*mock_context_, setProperty(testing::_, testing::_))
+        .WillByDefault(
+            [&](std::string_view, std::string_view) { return WasmResult::Ok; });
+  }
+  ~ModelMapperTest() override {}
+  std::unique_ptr<WasmBase> wasm_base_;
+  std::unique_ptr<WasmVm> test_vm_;
+  std::unique_ptr<MockContext> mock_context_;
+  std::unique_ptr<PluginRootContext> root_context_;
+  std::unique_ptr<PluginContext> context_;
+  std::string route_name_;
+  std::string service_name_;
+  std::string path_;
+  BufferBase body_;
+  BufferBase config_;
+};
+
+TEST_F(ModelMapperTest, ModelMappingTest) {
+  std::string configuration = R"(
+{
+  "modelMapping": {
+     "*": "qwen-long",
+     "gpt-4*": "qwen-max",
+     "gpt-4o": "qwen-turbo",
+     "gpt-4o-mini": "qwen-plus",
+     "text-embedding-v1": ""
+  }
+})";
+
+  config_.set(configuration);
+  EXPECT_TRUE(root_context_->configure(configuration.size()));
+
+  path_ = "/v1/chat/completions";
+  std::string request_json = R"({"model": "gpt-3.5"})";
+  EXPECT_CALL(*mock_context_,
+              setBuffer(testing::_, testing::_, testing::_, testing::_))
+      .WillOnce([&](WasmBufferType, size_t, size_t, std::string_view body) {
+        EXPECT_EQ(body, R"({"model":"qwen-long"})");
+        return WasmResult::Ok;
+      });
+
+  body_.set(request_json);
+  EXPECT_EQ(context_->onRequestHeaders(0, false),
+            FilterHeadersStatus::StopIteration);
+  EXPECT_EQ(context_->onRequestBody(20, true), FilterDataStatus::Continue);
+
+  request_json = R"({"model": "gpt-4"})";
+  EXPECT_CALL(*mock_context_,
+              setBuffer(testing::_, testing::_, testing::_, testing::_))
+      .WillOnce([&](WasmBufferType, size_t, size_t, std::string_view body) {
+        EXPECT_EQ(body, R"({"model":"qwen-max"})");
+        return WasmResult::Ok;
+      });
+
+  body_.set(request_json);
+  EXPECT_EQ(context_->onRequestHeaders(0, false),
+            FilterHeadersStatus::StopIteration);
+  EXPECT_EQ(context_->onRequestBody(20, true), FilterDataStatus::Continue);
+
+  request_json = R"({"model": "gpt-4o"})";
+  EXPECT_CALL(*mock_context_,
+              setBuffer(testing::_, testing::_, testing::_, testing::_))
+      .WillOnce([&](WasmBufferType, size_t, size_t, std::string_view body) {
+        EXPECT_EQ(body, R"({"model":"qwen-turbo"})");
+        return WasmResult::Ok;
+      });
+
+  body_.set(request_json);
+  EXPECT_EQ(context_->onRequestHeaders(0, false),
+            FilterHeadersStatus::StopIteration);
+  EXPECT_EQ(context_->onRequestBody(20, true), FilterDataStatus::Continue);
+
+  request_json = R"({"model": "gpt-4o-mini"})";
+  EXPECT_CALL(*mock_context_,
+              setBuffer(testing::_, testing::_, testing::_, testing::_))
+      .WillOnce([&](WasmBufferType, size_t, size_t, std::string_view body) {
+        EXPECT_EQ(body, R"({"model":"qwen-plus"})");
+        return WasmResult::Ok;
+      });
+
+  body_.set(request_json);
+  EXPECT_EQ(context_->onRequestHeaders(0, false),
+            FilterHeadersStatus::StopIteration);
+  EXPECT_EQ(context_->onRequestBody(20, true), FilterDataStatus::Continue);
+
+  request_json = R"({"model": "text-embedding-v1"})";
+  EXPECT_CALL(*mock_context_,
+              setBuffer(testing::_, testing::_, testing::_, testing::_))
+      .Times(0);
+
+  body_.set(request_json);
+  EXPECT_EQ(context_->onRequestHeaders(0, false),
+            FilterHeadersStatus::StopIteration);
+  EXPECT_EQ(context_->onRequestBody(20, true), FilterDataStatus::Continue);
+
+  request_json = R"({})";
+  EXPECT_CALL(*mock_context_,
+              setBuffer(testing::_, testing::_, testing::_, testing::_))
+      .WillOnce([&](WasmBufferType, size_t, size_t, std::string_view body) {
+        EXPECT_EQ(body, R"({"model":"qwen-long"})");
+        return WasmResult::Ok;
+      });
+
+  body_.set(request_json);
+  EXPECT_EQ(context_->onRequestHeaders(0, false),
+            FilterHeadersStatus::StopIteration);
+  EXPECT_EQ(context_->onRequestBody(20, true), FilterDataStatus::Continue);
+}
+
+TEST_F(ModelMapperTest, RouteLevelModelMappingTest) {
+  std::string configuration = R"(
+{
+  "_rules_": [
+    {
+      "_match_route_": ["route-a"],
+      "_match_service_": ["service-1"],
+      "modelMapping": {
+        "*": "qwen-long"
+      }
+    },
+    {
+      "_match_route_": ["route-b"],
+      "_match_service_": ["service-2"],
+      "modelMapping": {
+        "*": "qwen-max"
+      }
+    },
+    {
+      "_match_route_": ["route-b"],
+      "_match_service_": ["service-3"],
+      "modelMapping": {
+        "*": "qwen-turbo"
+      }
+    }
+]})";
+
+  config_.set(configuration);
+  EXPECT_TRUE(root_context_->configure(configuration.size()));
+
+  path_ = "/api/v1/chat/completions";
+  std::string request_json = R"({"model": "gpt-4"})";
+  body_.set(request_json);
+  route_name_ = "route-a";
+  service_name_ = "outbound|80||service-1";
+  EXPECT_CALL(*mock_context_,
+              setBuffer(testing::_, testing::_, testing::_, testing::_))
+      .WillOnce([&](WasmBufferType, size_t, size_t, std::string_view body) {
+        EXPECT_EQ(body, R"({"model":"qwen-long"})");
+        return WasmResult::Ok;
+      });
+  EXPECT_EQ(context_->onRequestHeaders(0, false),
+            FilterHeadersStatus::StopIteration);
+  EXPECT_EQ(context_->onRequestBody(20, true), FilterDataStatus::Continue);
+
+  route_name_ = "route-b";
+  service_name_ = "outbound|80||service-2";
+  EXPECT_CALL(*mock_context_,
+              setBuffer(testing::_, testing::_, testing::_, testing::_))
+      .WillOnce([&](WasmBufferType, size_t, size_t, std::string_view body) {
+        EXPECT_EQ(body, R"({"model":"qwen-max"})");
+        return WasmResult::Ok;
+      });
+  EXPECT_EQ(context_->onRequestHeaders(0, false),
+            FilterHeadersStatus::StopIteration);
+  EXPECT_EQ(context_->onRequestBody(20, true), FilterDataStatus::Continue);
+
+  route_name_ = "route-b";
+  service_name_ = "outbound|80||service-3";
+  EXPECT_CALL(*mock_context_,
+              setBuffer(testing::_, testing::_, testing::_, testing::_))
+      .WillOnce([&](WasmBufferType, size_t, size_t, std::string_view body) {
+        EXPECT_EQ(body, R"({"model":"qwen-turbo"})");
+        return WasmResult::Ok;
+      });
+  EXPECT_EQ(context_->onRequestHeaders(0, false),
+            FilterHeadersStatus::StopIteration);
+  EXPECT_EQ(context_->onRequestBody(20, true), FilterDataStatus::Continue);
+}
+
+}  // namespace model_mapper
+}  // namespace null_plugin
+}  // namespace proxy_wasm
--- a/plugins/wasm-cpp/extensions/model_router/BUILD
+++ b/plugins/wasm-cpp/extensions/model_router/BUILD
@@ -12,10 +12,10 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.

-load("@proxy_wasm_cpp_sdk//bazel/wasm:wasm.bzl", "wasm_cc_binary")
+load("@proxy_wasm_cpp_sdk//bazel:defs.bzl", "proxy_wasm_cc_binary")
 load("//bazel:wasm.bzl", "declare_wasm_image_targets")

-wasm_cc_binary(
+proxy_wasm_cc_binary(
    name = "model_router.wasm",
    srcs = [
        "plugin.cc",
--- a/plugins/wasm-cpp/extensions/model_router/README.md
+++ b/plugins/wasm-cpp/extensions/model_router/README.md
@@ -1,33 +1,35 @@
 ## 功能说明
 `model-router`插件实现了基于LLM协议中的model参数路由的功能

-## 运行属性
-
-插件执行阶段：`默认阶段`
-插件执行优先级：`260`
-
 ## 配置字段

-| 名称             | 数据类型        | 填写要求                | 默认值                 | 描述                                                  |
-| -----------      | --------------- | ----------------------- | ------                 | -------------------------------------------           |
-| `enable`         | bool            | 选填                    | false                  | 是否开启基于model参数路由                             |
-| `model_key`      | string          | 选填                    | model                  | 请求body中model参数的位置                             |
-| `add_header_key` | string          | 选填                    | x-higress-llm-provider | 从model参数中解析出的provider名字放到哪个请求header中 |
+| 名称                 | 数据类型        | 填写要求                | 默认值                   | 描述                                                  |
+| -----------          | --------------- | ----------------------- | ------                   | -------------------------------------------           |
+| `modelKey`           | string          | 选填                    | model                    | 请求body中model参数的位置                             |
+| `addProviderHeader`  | string          | 选填                    | -                        | 从model参数中解析出的provider名字放到哪个请求header中 |
+| `modelToHeader`      | string          | 选填                    | -                        | 直接将model参数放到哪个请求header中                   |
+| `enableOnPathSuffix` | array of string | 选填                    | ["/v1/chat/completions"] | 只对这些特定路径后缀的请求生效                        |

+## 运行属性
+
+插件执行阶段：认证阶段
+插件执行优先级：900

 ## 效果说明

-如下开启基于model参数路由的功能：
+### 基于 model 参数进行路由
+
+需要做如下配置：

 ```yaml
-enable: true
+modelToHeader: x-higress-llm-model
 ```

-开启后，插件将请求中 model 参数的 provider 部分（如果有）提取出来，设置到 x-higress-llm-provider 这个请求 header 中，用于后续路由，并将 model 参数重写为模型名称部分。举例来说，原生的 LLM 请求体是：
+插件会将请求中 model 参数提取出来，设置到 x-higress-llm-model 这个请求 header 中，用于后续路由，举例来说，原生的 LLM 请求体是：

 ```json
 {
-    "model": "qwen/qwen-long",
+    "model": "qwen-long",
    "frequency_penalty": 0,
    "max_tokens": 800,
    "stream": false,
@@ -43,7 +45,39 @@ enable: true

 经过这个插件后，将添加下面这个请求头(可以用于路由匹配)：

-x-higress-llm-provider: qwen
+x-higress-llm-model: qwen-long
+
+### 提取 model 参数中的 provider 字段用于路由
+
+> 注意这种模式需要客户端在 model 参数中通过`/`分隔的方式，来指定 provider
+
+需要做如下配置：
+
+```yaml
+addProviderHeader: x-higress-llm-provider
+```
+
+插件会将请求中 model 参数的 provider 部分（如果有）提取出来，设置到 x-higress-llm-provider 这个请求 header 中，用于后续路由，并将 model 参数重写为模型名称部分。举例来说，原生的 LLM 请求体是：
+
+```json
+{
+    "model": "dashscope/qwen-long",
+    "frequency_penalty": 0,
+    "max_tokens": 800,
+    "stream": false,
+    "messages": [{
+        "role": "user",
+        "content": "higress项目主仓库的github地址是什么"
+    }],
+    "presence_penalty": 0,
+    "temperature": 0.7,
+    "top_p": 0.95
+}
+```
+
+经过这个插件后，将添加下面这个请求头(可以用于路由匹配)：
+
+x-higress-llm-provider: dashscope

 原始的 LLM 请求体将被改成：

--- a/plugins/wasm-cpp/extensions/model_router/README_EN.md
+++ b/plugins/wasm-cpp/extensions/model_router/README_EN.md
@@ -1,38 +1,41 @@
 ## Function Description
-The `model-router` plugin implements the functionality of routing based on the `model` parameter in the LLM protocol.
-
-## Runtime Properties
-
-Plugin Execution Phase: `Default Phase`
-Plugin Execution Priority: `260`
+The `model-router` plugin implements the function of routing based on the model parameter in the LLM protocol.

 ## Configuration Fields

-| Name               | Data Type   | Filling Requirement | Default Value         | Description                                           |
-| -------------------- | ------------- | --------------------- | ---------------------- | ----------------------------------------------------- |
-| `enable`            | bool        | Optional             | false                 | Whether to enable routing based on the `model` parameter |
-| `model_key`         | string      | Optional             | model                 | The location of the `model` parameter in the request body |
-| `add_header_key`    | string      | Optional             | x-higress-llm-provider | The header where the parsed provider name from the `model` parameter will be placed |
+| Name                 | Data Type        | Filling Requirement                | Default Value                   | Description                                                  |
+| -----------          | --------------- | -----------------------            | ------                          | -------------------------------------------                  |
+| `modelKey`           | string           | Optional                           | model                           | The location of the model parameter in the request body       |
+| `addProviderHeader`  | string           | Optional                           | -                               | Which request header to place the provider name parsed from the model parameter |
+| `modelToHeader`      | string           | Optional                           | -                               | Which request header to directly place the model parameter    |
+| `enableOnPathSuffix` | array of string  | Optional                           | ["/v1/chat/completions"]        | Only effective for requests with these specific path suffixes |
+
+## Runtime Attributes
+
+Plugin execution phase: Authentication phase
+Plugin execution priority: 900

 ## Effect Description

-To enable routing based on the `model` parameter, use the following configuration:
+### Routing Based on the model Parameter
+
+The following configuration is required:

 ```yaml
-enable: true
+modelToHeader: x-higress-llm-model
 ```

-After enabling, the plugin extracts the provider part (if any) from the `model` parameter in the request, and sets it in the `x-higress-llm-provider` request header for subsequent routing. It also rewrites the `model` parameter to the model name part. For example, the original LLM request body is:
+The plugin will extract the model parameter from the request and set it in the x-higress-llm-model request header, which can be used for subsequent routing. For example, the original LLM request body:

 ```json
 {
-    "model": "openai/gpt-4o",
+    "model": "qwen-long",
    "frequency_penalty": 0,
    "max_tokens": 800,
    "stream": false,
    "messages": [{
        "role": "user",
-        "content": "What is the GitHub address for the main repository of the Higress project?"
+        "content": "What is the GitHub address of the main repository for the higress project"
    }],
    "presence_penalty": 0,
    "temperature": 0.7,
@@ -40,24 +43,55 @@ After enabling, the plugin extracts the provider part (if any) from the `model`
 }
 ```

-After processing by the plugin, the following request header (which can be used for routing matching) will be added:
+After processing by this plugin, the following request header (which can be used for route matching) will be added:

-`x-higress-llm-provider: openai`
+x-higress-llm-model: qwen-long

-The original LLM request body will be modified to:
+### Extracting the provider Field from the model Parameter for Routing
+
+> Note that this mode requires the client to specify the provider using a `/` separator in the model parameter.
+
+The following configuration is required:
+
+```yaml
+addProviderHeader: x-higress-llm-provider
+```
+
+The plugin will extract the provider part (if present) from the model parameter in the request and set it in the x-higress-llm-provider request header, which can be used for subsequent routing, and rewrite the model parameter to the model name part. For example, the original LLM request body:

 ```json
 {
-    "model": "gpt-4o",
+    "model": "dashscope/qwen-long",
    "frequency_penalty": 0,
    "max_tokens": 800,
    "stream": false,
    "messages": [{
        "role": "user",
-        "content": "What is the GitHub address for the main repository of the Higress project?"
+        "content": "What is the GitHub address of the main repository for the higress project"
    }],
    "presence_penalty": 0,
    "temperature": 0.7,
    "top_p": 0.95
 }
 ```
+
+After processing by this plugin, the following request header (which can be used for route matching) will be added:
+
+x-higress-llm-provider: dashscope
+
+The original LLM request body will be changed to:
+
+```json
+{
+    "model": "qwen-long",
+    "frequency_penalty": 0,
+    "max_tokens": 800,
+    "stream": false,
+    "messages": [{
+        "role": "user",
+        "content": "What is the GitHub address of the main repository for the higress project"
+    }],
+    "presence_penalty": 0,
+    "temperature": 0.7,
+    "top_p": 0.95
+}
--- a/plugins/wasm-cpp/extensions/model_router/plugin.cc
+++ b/plugins/wasm-cpp/extensions/model_router/plugin.cc
@@ -51,41 +51,54 @@ constexpr std::string_view DefaultMaxBodyBytes = "10485760";

 bool PluginRootContext::parsePluginConfig(const json& configuration,
                                          ModelRouterConfigRule& rule) {
-  if (auto it = configuration.find("enable"); it != configuration.end()) {
-    if (it->is_boolean()) {
-      rule.enable_ = it->get<bool>();
-    } else {
-      LOG_WARN("Invalid type for enable. Expected boolean.");
-      return false;
-    }
-  }
-
-  if (auto it = configuration.find("model_key"); it != configuration.end()) {
+  if (auto it = configuration.find("modelKey"); it != configuration.end()) {
    if (it->is_string()) {
      rule.model_key_ = it->get<std::string>();
    } else {
-      LOG_WARN("Invalid type for model_key. Expected string.");
+      LOG_ERROR("Invalid type for modelKey. Expected string.");
      return false;
    }
  }

-  if (auto it = configuration.find("add_header_key");
+  if (auto it = configuration.find("addProviderHeader");
      it != configuration.end()) {
    if (it->is_string()) {
-      rule.add_header_key_ = it->get<std::string>();
+      rule.add_provider_header_ = it->get<std::string>();
    } else {
-      LOG_WARN("Invalid type for add_header_key. Expected string.");
+      LOG_ERROR("Invalid type for addProviderHeader. Expected string.");
      return false;
    }
  }

+  if (auto it = configuration.find("modelToHeader");
+      it != configuration.end()) {
+    if (it->is_string()) {
+      rule.model_to_header_ = it->get<std::string>();
+    } else {
+      LOG_ERROR("Invalid type for modelToHeader. Expected string.");
+      return false;
+    }
+  }
+
+  if (!JsonArrayIterate(
+          configuration, "enableOnPathSuffix", [&](const json& item) -> bool {
+            if (item.is_string()) {
+              rule.enable_on_path_suffix_.emplace_back(item.get<std::string>());
+              return true;
+            }
+            return false;
+          })) {
+    LOG_ERROR("Invalid type for item in enableOnPathSuffix. Expected string.");
+    return false;
+  }
+
  return true;
 }

 bool PluginRootContext::onConfigure(size_t size) {
  // Parse configuration JSON string.
  if (size > 0 && !configure(size)) {
-    LOG_WARN("configuration has errors initialization will not continue.");
+    LOG_ERROR("configuration has errors initialization will not continue.");
    return false;
  }
  return true;
@@ -97,13 +110,13 @@ bool PluginRootContext::configure(size_t configuration_size) {
  // Parse configuration JSON string.
  auto result = ::Wasm::Common::JsonParse(configuration_data->view());
  if (!result) {
-    LOG_WARN(absl::StrCat("cannot parse plugin configuration JSON string: ",
-                          configuration_data->view()));
+    LOG_ERROR(absl::StrCat("cannot parse plugin configuration JSON string: ",
+                           configuration_data->view()));
    return false;
  }
  if (!parseRuleConfig(result.value())) {
-    LOG_WARN(absl::StrCat("cannot parse plugin configuration JSON string: ",
-                          configuration_data->view()));
+    LOG_ERROR(absl::StrCat("cannot parse plugin configuration JSON string: ",
+                           configuration_data->view()));
    return false;
  }
  return true;
@@ -111,7 +124,25 @@ bool PluginRootContext::configure(size_t configuration_size) {

 FilterHeadersStatus PluginRootContext::onHeader(
    const ModelRouterConfigRule& rule) {
-  if (!rule.enable_ || !Wasm::Common::Http::hasRequestBody()) {
+  if (!Wasm::Common::Http::hasRequestBody()) {
+    return FilterHeadersStatus::Continue;
+  }
+  auto path = getRequestHeader(Wasm::Common::Http::Header::Path)->toString();
+  auto params_pos = path.find('?');
+  size_t uri_end;
+  if (params_pos == std::string::npos) {
+    uri_end = path.size();
+  } else {
+    uri_end = params_pos;
+  }
+  bool enable = false;
+  for (const auto& enable_suffix : rule.enable_on_path_suffix_) {
+    if (absl::EndsWith({path.c_str(), uri_end}, enable_suffix)) {
+      enable = true;
+      break;
+    }
+  }
+  if (!enable) {
    return FilterHeadersStatus::Continue;
  }
  auto content_type_value =
@@ -128,7 +159,8 @@ FilterHeadersStatus PluginRootContext::onHeader(
 FilterDataStatus PluginRootContext::onBody(const ModelRouterConfigRule& rule,
                                           std::string_view body) {
  const auto& model_key = rule.model_key_;
-  const auto& add_header_key = rule.add_header_key_;
+  const auto& add_provider_header = rule.add_provider_header_;
+  const auto& model_to_header = rule.model_to_header_;
  auto body_json_opt = ::Wasm::Common::JsonParse(body);
  if (!body_json_opt) {
    LOG_WARN(absl::StrCat("cannot parse body to JSON string: ", body));
@@ -137,18 +169,24 @@ FilterDataStatus PluginRootContext::onBody(const ModelRouterConfigRule& rule,
  auto body_json = body_json_opt.value();
  if (body_json.contains(model_key)) {
    std::string model_value = body_json[model_key];
-    auto pos = model_value.find('/');
-    if (pos != std::string::npos) {
-      const auto& provider = model_value.substr(0, pos);
-      const auto& model = model_value.substr(pos + 1);
-      replaceRequestHeader(add_header_key, provider);
-      body_json[model_key] = model;
-      setBuffer(WasmBufferType::HttpRequestBody, 0,
-                std::numeric_limits<size_t>::max(), body_json.dump());
-      LOG_DEBUG(absl::StrCat("model route to provider:", provider,
-                             ", model:", model));
-    } else {
-      LOG_DEBUG(absl::StrCat("model route not work, model:", model_value));
+    if (!model_to_header.empty()) {
+      replaceRequestHeader(model_to_header, model_value);
+    }
+    if (!add_provider_header.empty()) {
+      auto pos = model_value.find('/');
+      if (pos != std::string::npos) {
+        const auto& provider = model_value.substr(0, pos);
+        const auto& model = model_value.substr(pos + 1);
+        replaceRequestHeader(add_provider_header, provider);
+        body_json[model_key] = model;
+        setBuffer(WasmBufferType::HttpRequestBody, 0,
+                  std::numeric_limits<size_t>::max(), body_json.dump());
+        LOG_DEBUG(absl::StrCat("model route to provider:", provider,
+                               ", model:", model));
+      } else {
+        LOG_DEBUG(absl::StrCat("model route to provider not work, model:",
+                               model_value));
+      }
    }
  }
  return FilterDataStatus::Continue;
--- a/plugins/wasm-cpp/extensions/model_router/plugin.h
+++ b/plugins/wasm-cpp/extensions/model_router/plugin.h
@@ -37,9 +37,10 @@ namespace model_router {
 #endif

 struct ModelRouterConfigRule {
-  bool enable_ = false;
  std::string model_key_ = "model";
-  std::string add_header_key_ = "x-higress-llm-provider";
+  std::string add_provider_header_;
+  std::string model_to_header_;
+  std::vector<std::string> enable_on_path_suffix_ = {"/v1/chat/completions"};
 };

 // PluginRootContext is the root context for all streams processed by the
--- a/plugins/wasm-cpp/extensions/model_router/plugin_test.cc
+++ b/plugins/wasm-cpp/extensions/model_router/plugin_test.cc
@@ -89,6 +89,8 @@ class ModelRouterTest : public ::testing::Test {
            *result = "application/json";
          } else if (header == "content-length") {
            *result = "1024";
+          } else if (header == ":path") {
+            *result = path_;
          }
          return WasmResult::Ok;
        });
@@ -122,6 +124,7 @@ class ModelRouterTest : public ::testing::Test {
  std::unique_ptr<PluginRootContext> root_context_;
  std::unique_ptr<PluginContext> context_;
  std::string route_name_;
+  std::string path_;
  BufferBase body_;
  BufferBase config_;
 };
@@ -129,12 +132,13 @@ class ModelRouterTest : public ::testing::Test {
 TEST_F(ModelRouterTest, RewriteModelAndHeader) {
  std::string configuration = R"(
 {
-  "enable": true
+  "addProviderHeader": "x-higress-llm-provider"
    })";

  config_.set(configuration);
  EXPECT_TRUE(root_context_->configure(configuration.size()));

+  path_ = "/v1/chat/completions";
  std::string request_json = R"({"model": "qwen/qwen-long"})";
  EXPECT_CALL(*mock_context_,
              setBuffer(testing::_, testing::_, testing::_, testing::_))
@@ -154,19 +158,73 @@ TEST_F(ModelRouterTest, RewriteModelAndHeader) {
  EXPECT_EQ(context_->onRequestBody(28, true), FilterDataStatus::Continue);
 }

+TEST_F(ModelRouterTest, ModelToHeader) {
+  std::string configuration = R"(
+{
+  "modelToHeader": "x-higress-llm-model"
+    })";
+
+  config_.set(configuration);
+  EXPECT_TRUE(root_context_->configure(configuration.size()));
+
+  path_ = "/v1/chat/completions";
+  std::string request_json = R"({"model": "qwen-long"})";
+  EXPECT_CALL(*mock_context_,
+              setBuffer(testing::_, testing::_, testing::_, testing::_))
+      .Times(0);
+
+  EXPECT_CALL(
+      *mock_context_,
+      replaceHeaderMapValue(testing::_, std::string_view("x-higress-llm-model"),
+                            std::string_view("qwen-long")));
+
+  body_.set(request_json);
+  EXPECT_EQ(context_->onRequestHeaders(0, false),
+            FilterHeadersStatus::StopIteration);
+  EXPECT_EQ(context_->onRequestBody(28, true), FilterDataStatus::Continue);
+}
+
+TEST_F(ModelRouterTest, IgnorePath) {
+  std::string configuration = R"(
+{
+  "addProviderHeader": "x-higress-llm-provider"
+    })";
+
+  config_.set(configuration);
+  EXPECT_TRUE(root_context_->configure(configuration.size()));
+
+  path_ = "/v1/chat/xxxx";
+  std::string request_json = R"({"model": "qwen/qwen-long"})";
+  EXPECT_CALL(*mock_context_,
+              setBuffer(testing::_, testing::_, testing::_, testing::_))
+      .Times(0);
+
+  EXPECT_CALL(*mock_context_,
+              replaceHeaderMapValue(testing::_,
+                                    std::string_view("x-higress-llm-provider"),
+                                    std::string_view("qwen")))
+      .Times(0);
+
+  body_.set(request_json);
+  EXPECT_EQ(context_->onRequestHeaders(0, false),
+            FilterHeadersStatus::Continue);
+  EXPECT_EQ(context_->onRequestBody(28, true), FilterDataStatus::Continue);
+}
+
 TEST_F(ModelRouterTest, RouteLevelRewriteModelAndHeader) {
  std::string configuration = R"(
 {
  "_rules_": [
    {
      "_match_route_": ["route-a"],
-      "enable": true
+      "addProviderHeader": "x-higress-llm-provider"
    }
 ]})";

  config_.set(configuration);
  EXPECT_TRUE(root_context_->configure(configuration.size()));

+  path_ = "/api/v1/chat/completions";
  std::string request_json = R"({"model": "qwen/qwen-long"})";
  EXPECT_CALL(*mock_context_,
              setBuffer(testing::_, testing::_, testing::_, testing::_))
--- a/plugins/wasm-cpp/extensions/request_block/BUILD
+++ b/plugins/wasm-cpp/extensions/request_block/BUILD
@@ -12,10 +12,10 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.

-load("@proxy_wasm_cpp_sdk//bazel/wasm:wasm.bzl", "wasm_cc_binary")
+load("@proxy_wasm_cpp_sdk//bazel:defs.bzl", "proxy_wasm_cc_binary")
 load("//bazel:wasm.bzl", "declare_wasm_image_targets")

-wasm_cc_binary(
+proxy_wasm_cc_binary(
    name = "request_block.wasm",
    srcs = [
        "plugin.cc",
@@ -27,7 +27,6 @@ wasm_cc_binary(
        "//common:json_util",
        "//common:http_util",
        "//common:rule_util",
-        "@proxy_wasm_cpp_sdk//:proxy_wasm_intrinsics",
    ],
 )

--- a/plugins/wasm-cpp/extensions/sni_misdirect/BUILD
+++ b/plugins/wasm-cpp/extensions/sni_misdirect/BUILD
@@ -12,10 +12,10 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.

-load("@proxy_wasm_cpp_sdk//bazel/wasm:wasm.bzl", "wasm_cc_binary")
+load("@proxy_wasm_cpp_sdk//bazel:defs.bzl", "proxy_wasm_cc_binary")
 load("//bazel:wasm.bzl", "declare_wasm_image_targets")

-wasm_cc_binary(
+proxy_wasm_cc_binary(
    name = "sni_misdirect.wasm",
    srcs = [
        "plugin.cc",
@@ -25,7 +25,6 @@ wasm_cc_binary(
        "@com_google_absl//absl/strings:str_format",        
        "@com_google_absl//absl/strings",
        "//common:http_util",
-        "@proxy_wasm_cpp_sdk//:proxy_wasm_intrinsics",
    ],
 )

--- a/plugins/wasm-go/examples/custom-log/config.yaml
+++ b/plugins/wasm-go/examples/custom-log/config.yaml
@@ -0,0 +1,68 @@
+static_resources:
+  listeners:
+  - name: listener_0
+    address:
+      socket_address:
+        protocol: TCP
+        address: 0.0.0.0
+        port_value: 8080
+    filter_chains:
+    - filters:
+      - name: envoy.filters.network.http_connection_manager
+        typed_config:
+          "@type": type.googleapis.com/envoy.extensions.filters.network.http_connection_manager.v3.HttpConnectionManager
+          stat_prefix: ingress_http
+          access_log:
+          - name: envoy.access_loggers.file
+            typed_config:
+              "@type": type.googleapis.com/envoy.extensions.access_loggers.file.v3.FileAccessLog
+              log_format:
+                text_format_source:
+                  inline_string: "{\"custom_log\":\"%FILTER_STATE(wasm.custom_log:PLAIN)%\",\"ai_log\":\"%FILTER_STATE(wasm.ai_log:PLAIN)%\"}
+                  
+                  "
+              path: /dev/stdout
+          route_config:
+            name: local_route
+            virtual_hosts:
+            - name: local_service
+              domains: ["*"]
+              routes:
+              - name: get
+                match:
+                  prefix: "/get"
+                route:
+                  cluster: httpbin
+          http_filters:
+          - name: test
+            typed_config:
+              "@type": type.googleapis.com/udpa.type.v1.TypedStruct
+              type_url: type.googleapis.com/envoy.extensions.filters.http.wasm.v3.Wasm
+              value:
+                config:
+                  name: test
+                  vm_config:
+                    runtime: envoy.wasm.runtime.v8
+                    code:
+                      local:
+                        filename: main.wasm
+                  configuration:
+                    "@type": "type.googleapis.com/google.protobuf.StringValue"
+                    value: {}
+          - name: envoy.filters.http.router
+            typed_config:
+              "@type": type.googleapis.com/envoy.extensions.filters.http.router.v3.Router
+  clusters:
+  - name: httpbin
+    connect_timeout: 600s
+    type: STRICT_DNS
+    lb_policy: ROUND_ROBIN
+    load_assignment:
+      cluster_name: httpbin
+      endpoints:
+        - lb_endpoints:
+          - endpoint:
+              address:
+                socket_address:
+                  address: httpbin.org
+                  port_value: 80
--- a/plugins/wasm-go/examples/custom-log/go.mod
+++ b/plugins/wasm-go/examples/custom-log/go.mod
@@ -0,0 +1,20 @@
+module github.com/alibaba/higress/plugins/wasm-go/extensions/custom-logs
+
+go 1.18
+
+replace github.com/alibaba/higress/plugins/wasm-go => ../..
+
+require (
+	github.com/alibaba/higress/plugins/wasm-go v0.0.0
+	github.com/higress-group/proxy-wasm-go-sdk v1.0.0
+)
+
+require (
+	github.com/google/uuid v1.3.0 // indirect
+	github.com/higress-group/nottinygc v0.0.0-20231101025119-e93c4c2f8520 // indirect
+	github.com/magefile/mage v1.14.0 // indirect
+	github.com/tidwall/gjson v1.17.3 // indirect
+	github.com/tidwall/match v1.1.1 // indirect
+	github.com/tidwall/pretty v1.2.0 // indirect
+	github.com/tidwall/resp v0.1.1 // indirect
+)
--- a/plugins/wasm-go/examples/custom-log/go.sum
+++ b/plugins/wasm-go/examples/custom-log/go.sum
@@ -0,0 +1,20 @@
+github.com/davecgh/go-spew v1.1.1 h1:vj9j/u1bqnvCEfJOwUhtlOARqs3+rkHYY13jYWTU97c=
+github.com/google/uuid v1.3.0 h1:t6JiXgmwXMjEs8VusXIJk2BXHsn+wx8BZdTaoZ5fu7I=
+github.com/google/uuid v1.3.0/go.mod h1:TIyPZe4MgqvfeYDBFedMoGGpEw/LqOeaOT+nhxU+yHo=
+github.com/higress-group/nottinygc v0.0.0-20231101025119-e93c4c2f8520 h1:IHDghbGQ2DTIXHBHxWfqCYQW1fKjyJ/I7W1pMyUDeEA=
+github.com/higress-group/nottinygc v0.0.0-20231101025119-e93c4c2f8520/go.mod h1:Nz8ORLaFiLWotg6GeKlJMhv8cci8mM43uEnLA5t8iew=
+github.com/higress-group/proxy-wasm-go-sdk v1.0.0 h1:BZRNf4R7jr9hwRivg/E29nkVaKEak5MWjBDhWjuHijU=
+github.com/higress-group/proxy-wasm-go-sdk v1.0.0/go.mod h1:iiSyFbo+rAtbtGt/bsefv8GU57h9CCLYGJA74/tF5/0=
+github.com/magefile/mage v1.14.0 h1:6QDX3g6z1YvJ4olPhT1wksUcSa/V0a1B+pJb73fBjyo=
+github.com/magefile/mage v1.14.0/go.mod h1:z5UZb/iS3GoOSn0JgWuiw7dxlurVYTu+/jHXqQg881A=
+github.com/pmezard/go-difflib v1.0.0 h1:4DBwDE0NGyQoBHbLQYPwSUPoCMWR5BEzIk/f1lZbAQM=
+github.com/stretchr/testify v1.8.4 h1:CcVxjf3Q8PM0mHUKJCdn+eZZtm5yQwehR5yeSVQQcUk=
+github.com/tidwall/gjson v1.17.3 h1:bwWLZU7icoKRG+C+0PNwIKC6FCJO/Q3p2pZvuP0jN94=
+github.com/tidwall/gjson v1.17.3/go.mod h1:/wbyibRr2FHMks5tjHJ5F8dMZh3AcwJEMf5vlfC0lxk=
+github.com/tidwall/match v1.1.1 h1:+Ho715JplO36QYgwN9PGYNhgZvoUSc9X2c80KVTi+GA=
+github.com/tidwall/match v1.1.1/go.mod h1:eRSPERbgtNPcGhD8UCthc6PmLEQXEWd3PRB5JTxsfmM=
+github.com/tidwall/pretty v1.2.0 h1:RWIZEg2iJ8/g6fDDYzMpobmaoGh5OLl4AXtGUGPcqCs=
+github.com/tidwall/pretty v1.2.0/go.mod h1:ITEVvHYasfjBbM0u2Pg8T2nJnzm8xPwvNhhsoaGGjNU=
+github.com/tidwall/resp v0.1.1 h1:Ly20wkhqKTmDUPlyM1S7pWo5kk0tDu8OoC/vFArXmwE=
+github.com/tidwall/resp v0.1.1/go.mod h1:3/FrruOBAxPTPtundW0VXgmsQ4ZBA0Aw714lVYgwFa0=
+gopkg.in/yaml.v3 v3.0.1 h1:fxVm/GzAzEWqLHuvctI91KS9hhNmmWOoWu0XTYJS7CA=
--- a/plugins/wasm-go/examples/custom-log/main.go
+++ b/plugins/wasm-go/examples/custom-log/main.go
@@ -0,0 +1,67 @@
+// Copyright (c) 2022 Alibaba Group Holding Ltd.
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+//      http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+package main
+
+import (
+	"math/rand"
+
+	"github.com/higress-group/proxy-wasm-go-sdk/proxywasm/types"
+
+	"github.com/alibaba/higress/plugins/wasm-go/pkg/wrapper"
+)
+
+func main() {
+	wrapper.SetCtx(
+		"custom-log",
+		wrapper.ProcessRequestHeadersBy(onHttpRequestHeaders),
+	)
+}
+
+type CustomLogConfig struct {
+}
+
+// Method 1: write custom log
+func writeLog(ctx wrapper.HttpContext) {
+	ctx.SetUserAttribute("question", "当然可以。在Python中，你可以创建一个函数来计算一系列数字的和。下面是一个简单的例子，该函数接受一个数字列表作为输入，并返回它们的总和。\n\n```python\ndef sum_of_numbers(numbers):\n    \"\"\"\n    计算列表中所有数字的和。\n    \n    参数:\n    numbers (list of int or float): 一个包含数字的列表。\n    \n    返回:\n    int or float: 列表中所有数字的总和。\n    \"\"\"\n    total_sum = sum(numbers)  # 使用Python内置的sum函数计算总和\n    return total_sum\n\n# 示例使用\nnumbers_list = [1, 2, 3, 4, 5]\nprint(\"The sum is:\", sum_of_numbers(numbers_list))  # 输出：The sum is: 15\n```\n\n在这段代码中，我们定义了一个名为 `sum_of_numbers` 的函数，它接收一个参数 `numbers`，这是一个包含整数或浮点数的列表。函数内部使用了Python的内置函数 `sum()` 来计算这些数字的总和，并将结果返回。\n\n你也可以手动实现求和逻辑，而不是使用内置的 `sum()` 函数，如下所示：\n\n```python\ndef sum_of_numbers_manual(numbers):\n    \"\"\"\n    手动计算列表中所有数字的和。\n    \n    参数:\n    numbers (list of int or float): 一个包含数字的列表。\n    \n    返回:\n    int or float: 列表中所有数字的总和。\n    \"\"\"\n    total_sum = 0\n    for number in numbers:\n        total_sum += number\n    return total_sum\n\n# 示例使用\nnumbers_list = [1, 2, 3, 4, 5]\nprint(\"The sum is:\", sum_of_numbers_manual(numbers_list))  # 输出：The sum is: 15\n```\n\n在这个版本中，我们初始化 `total_sum` 为0，然后遍历列表中的每个元素，并将其加到 `total_sum` 上。最后返回这个累加的结果。这两种方法都可以达到相同的目的，但是使用内置函数通常更简洁且效率更高。")
+	ctx.SetUserAttribute("k2", 2213.22)
+	ctx.WriteUserAttributeToLog()
+}
+
+// Methods 2: write custom log with specific key
+func writeLogWithKey(ctx wrapper.HttpContext, key string) {
+	ctx.SetUserAttribute("k2", 2213.22)
+	_ = ctx.WriteUserAttributeToLogWithKey(key)
+	ctx.SetUserAttribute("k2", 212939.22)
+	ctx.SetUserAttribute("k3", 123)
+	_ = ctx.WriteUserAttributeToLogWithKey(key)
+}
+
+// Methods 2: write custom log with specific key
+func writeTraceAttribute(ctx wrapper.HttpContext) {
+	ctx.SetUserAttribute("question", "当然可以。在Python中，你可以创建一个函数来计算一系列数字的和。下面是一个简单的例子，该函数接受一个数字列表作为输入，并返回它们的总和。\n\n```python\ndef sum_of_numbers(numbers):\n    \"\"\"\n    计算列表中所有数字的和。\n    \n    参数:\n    numbers (list of int or float): 一个包含数字的列表。\n    \n    返回:\n    int or float: 列表中所有数字的总和。\n    \"\"\"\n    total_sum = sum(numbers)  # 使用Python内置的sum函数计算总和\n    return total_sum\n\n# 示例使用\nnumbers_list = [1, 2, 3, 4, 5]\nprint(\"The sum is:\", sum_of_numbers(numbers_list))  # 输出：The sum is: 15\n```\n\n在这段代码中，我们定义了一个名为 `sum_of_numbers` 的函数，它接收一个参数 `numbers`，这是一个包含整数或浮点数的列表。函数内部使用了Python的内置函数 `sum()` 来计算这些数字的总和，并将结果返回。\n\n你也可以手动实现求和逻辑，而不是使用内置的 `sum()` 函数，如下所示：\n\n```python\ndef sum_of_numbers_manual(numbers):\n    \"\"\"\n    手动计算列表中所有数字的和。\n    \n    参数:\n    numbers (list of int or float): 一个包含数字的列表。\n    \n    返回:\n    int or float: 列表中所有数字的总和。\n    \"\"\"\n    total_sum = 0\n    for number in numbers:\n        total_sum += number\n    return total_sum\n\n# 示例使用\nnumbers_list = [1, 2, 3, 4, 5]\nprint(\"The sum is:\", sum_of_numbers_manual(numbers_list))  # 输出：The sum is: 15\n```\n\n在这个版本中，我们初始化 `total_sum` 为0，然后遍历列表中的每个元素，并将其加到 `total_sum` 上。最后返回这个累加的结果。这两种方法都可以达到相同的目的，但是使用内置函数通常更简洁且效率更高。")
+	ctx.SetUserAttribute("k2", 2213.22)
+	ctx.WriteUserAttributeToTrace()
+}
+
+func onHttpRequestHeaders(ctx wrapper.HttpContext, config CustomLogConfig, log wrapper.Log) types.Action {
+	if rand.Intn(10)%3 == 1 {
+		writeLog(ctx)
+	} else if rand.Intn(10)%3 == 2 {
+		writeLogWithKey(ctx, "ai_log")
+	} else {
+		writeTraceAttribute(ctx)
+	}
+	return types.ActionContinue
+}
--- a/plugins/wasm-go/extensions/ai-agent/README.md
+++ b/plugins/wasm-go/extensions/ai-agent/README.md
@@ -5,7 +5,7 @@ description: AI Agent插件配置参考
 ---

 ## 功能说明
-一个可定制化的 API AI Agent，支持配置 http method 类型为 GET 与 POST 的 API，支持多轮对话，支持流式与非流式模式。
+一个可定制化的 API AI Agent，支持配置 http method 类型为 GET 与 POST 的 API，支持多轮对话，支持流式与非流式模式，支持将结果格式化为自定义的 json。
 agent流程图如下：
 ![ai-agent](https://img.alicdn.com/imgextra/i1/O1CN01PGSDW31WQfEPm173u_!!6000000002783-0-tps-2733-1473.jpg)

@@ -21,6 +21,7 @@ agent流程图如下：
 | `llm`            | object    | 必填    | -      | 配置 AI 服务提供商的信息     |
 | `apis`           | object    | 必填    | -      | 配置外部 API 服务提供商的信息 |
 | `promptTemplate` | object    | 非必填  | -      | 配置 Agent ReAct 模板的信息  |
+| `jsonResp`       | object    | 非必填  | -      | 配置 json 格式化的相关信息   |

 `llm`的配置字段说明如下：

@@ -78,7 +79,14 @@ agent流程图如下：
 | `observation`   | string    | 非必填     | -      | Agent ReAct 模板的 observation 部分          |
 | `thought2`      | string    | 非必填     | -      | Agent ReAct 模板的 thought2 部分             |

-## 用法示例
+`jsonResp`的配置字段说明如下：
+
+| 名称               | 数据类型   | 填写要求 | 默认值 | 描述                               |
+|--------------------|-----------|---------|--------|-----------------------------------|
+| `enable`           | bool      | 非必填   | false  | 是否开启 json 格式化。             |
+| `jsonSchema`       | string    | 非必填   | -      | 自定义 json schema                |
+
+## 用法示例-不开启 json 格式化

 **配置信息**

@@ -293,7 +301,7 @@ deepl提供了一个工具，用于翻译给定的句子，支持多语言。。
 **请求示例**

 ```shell
-curl 'http://<这里换成网关公网IP>/api/openai/v1/chat/completions' \
+curl 'http://<这里换成网关地址>/api/openai/v1/chat/completions' \
 -H 'Accept: application/json, text/event-stream' \
 -H 'Content-Type: application/json' \
 --data-raw '{"model":"qwen","frequency_penalty":0,"max_tokens":800,"stream":false,"messages":[{"role":"user","content":"我想在济南市鑫盛大厦附近喝咖啡，给我推荐几个"}],"presence_penalty":0,"temperature":0,"top_p":0}'
@@ -308,7 +316,7 @@ curl 'http://<这里换成网关公网IP>/api/openai/v1/chat/completions' \
 **请求示例**

 ```shell
-curl 'http://<这里换成网关公网IP>/api/openai/v1/chat/completions' \
+curl 'http://<这里换成网关地址>/api/openai/v1/chat/completions' \
 -H 'Accept: application/json, text/event-stream' \
 -H 'Content-Type: application/json' \
 --data-raw '{"model":"qwen","frequency_penalty":0,"max_tokens":800,"stream":false,"messages":[{"role":"user","content":"济南市现在的天气情况如何？"}],"presence_penalty":0,"temperature":0,"top_p":0}'
@@ -323,7 +331,7 @@ curl 'http://<这里换成网关公网IP>/api/openai/v1/chat/completions' \
 **请求示例**

 ```shell
-curl 'http://<这里换成网关公网IP>/api/openai/v1/chat/completions' \
+curl 'http://<这里换成网关地址>/api/openai/v1/chat/completions' \
 -H 'Accept: application/json, text/event-stream' \
 -H 'Content-Type: application/json' \
 --data-raw '{"model":"qwen","frequency_penalty":0,"max_tokens":800,"stream":false,"messages":[{"role": "user","content": "济南的天气如何？"},{ "role": "assistant","content": "目前，济南市的天气为多云，气温为24℃，数据更新时间为2024年9月12日21时50分14秒。"},{"role": "user","content": "北京呢？"}],"presence_penalty":0,"temperature":0,"top_p":0}'
@@ -338,7 +346,7 @@ curl 'http://<这里换成网关公网IP>/api/openai/v1/chat/completions' \
 **请求示例**

 ```shell
-curl 'http://<这里换成网关公网IP>/api/openai/v1/chat/completions' \
+curl 'http://<这里换成网关地址>/api/openai/v1/chat/completions' \
 -H 'Accept: application/json, text/event-stream' \
 -H 'Content-Type: application/json' \
 --data-raw '{"model":"qwen","frequency_penalty":0,"max_tokens":800,"stream":false,"messages":[{"role":"user","content":"济南市现在的天气情况如何？用华氏度表示，用日语回答"}],"presence_penalty":0,"temperature":0,"top_p":0}'
@@ -353,7 +361,7 @@ curl 'http://<这里换成网关公网IP>/api/openai/v1/chat/completions' \
 **请求示例**

 ```shell
-curl 'http://<这里换成网关公网IP>/api/openai/v1/chat/completions' \
+curl 'http://<这里换成网关地址>/api/openai/v1/chat/completions' \
 -H 'Accept: application/json, text/event-stream' \
 -H 'Content-Type: application/json' \
 --data-raw '{"model":"qwen","frequency_penalty":0,"max_tokens":800,"stream":false,"messages":[{"role":"user","content":"帮我用德语翻译以下句子：九头蛇万岁!"}],"presence_penalty":0,"temperature":0,"top_p":0}'
@@ -364,3 +372,71 @@ curl 'http://<这里换成网关公网IP>/api/openai/v1/chat/completions' \
 ```json
 {"id":"65dcf12c-61ff-9e68-bffa-44fc9e6070d5","choices":[{"index":0,"message":{"role":"assistant","content":" “九头蛇万岁!”的德语翻译为“Hoch lebe Hydra!”。"},"finish_reason":"stop"}],"created":1724043865,"model":"qwen-max-0403","object":"chat.completion","usage":{"prompt_tokens":908,"completion_tokens":52,"total_tokens":960}}
 ```
+
+## 用法示例-开启 json 格式化
+
+**配置信息**
+在上述配置的基础上增加 jsonResp 配置
+```yaml
+jsonResp:
+  enable: true
+```
+
+**请求示例**
+
+```shell
+curl 'http://<这里换成网关地址>/api/openai/v1/chat/completions' \
+-H 'Accept: application/json, text/event-stream' \
+-H 'Content-Type: application/json' \
+--data-raw '{"model":"qwen","frequency_penalty":0,"max_tokens":800,"stream":false,"messages":[{"role":"user","content":"北京市现在的天气情况如何？"}],"presence_penalty":0,"temperature":0,"top_p":0}'
+```
+
+**响应示例**
+
+```json
+{"id":"ebd6ea91-8e38-9e14-9a5b-90178d2edea4","choices":[{"index":0,"message":{"role":"assistant","content": "{\"city\": \"北京市\", \"weather_condition\": \"多云\", \"temperature\": \"19℃\", \"data_update_time\": \"2024年10月9日16时37分53秒\"}"},"finish_reason":"stop"}],"created":1723187991,"model":"qwen-max-0403","object":"chat.completion","usage":{"prompt_tokens":890,"completion_tokens":56,"total_tokens":946}}
+```
+如果不自定义 json schema，大模型会自动生成一个 json 格式
+
+**配置信息**
+增加自定义 json schema 配置
+```yaml
+jsonResp:
+  enable: true
+  jsonSchema: |
+    title: WeatherSchema
+    type: object
+    properties:
+      location:
+        type: string
+        description: 城市名称.
+      weather:
+        type: string
+        description: 天气情况.
+      temperature:
+        type: string
+        description: 温度.
+      update_time:
+        type: string
+        description: 数据更新时间.
+    required:
+      - location
+      - weather
+      - temperature
+    additionalProperties: false
+```
+
+**请求示例**
+
+```shell
+curl 'http://<这里换成网关地址>/api/openai/v1/chat/completions' \
+-H 'Accept: application/json, text/event-stream' \
+-H 'Content-Type: application/json' \
+--data-raw '{"model":"qwen","frequency_penalty":0,"max_tokens":800,"stream":false,"messages":[{"role":"user","content":"北京市现在的天气情况如何？"}],"presence_penalty":0,"temperature":0,"top_p":0}'
+```
+
+**响应示例**
+
+```json
+{"id":"ebd6ea91-8e38-9e14-9a5b-90178d2edea4","choices":[{"index":0,"message":{"role":"assistant","content": "{\"location\": \"北京市\", \"weather\": \"多云\", \"temperature\": \"19℃\", \"update_time\": \"2024年10月9日16时37分53秒\"}"},"finish_reason":"stop"}],"created":1723187991,"model":"qwen-max-0403","object":"chat.completion","usage":{"prompt_tokens":890,"completion_tokens":56,"total_tokens":946}}
+```
--- a/plugins/wasm-go/extensions/ai-agent/README_EN.md
+++ b/plugins/wasm-go/extensions/ai-agent/README_EN.md
@@ -4,7 +4,7 @@ keywords: [ AI Gateway, AI Agent ]
 description: AI Agent plugin configuration reference
 ---
 ## Functional Description
-A customizable API AI Agent that supports configuring HTTP method types as GET and POST APIs. Supports multiple dialogue rounds, streaming and non-streaming modes.  
+A customizable API AI Agent that supports configuring HTTP method types as GET and POST APIs. Supports multiple dialogue rounds, streaming and non-streaming modes, support for formatting results as custom json.  
 The agent flow chart is as follows:  
 ![ai-agent](https://github.com/user-attachments/assets/b0761a0c-1afa-496c-a98e-bb9f38b340f8)  

@@ -20,6 +20,7 @@ Plugin execution priority: `200`
 | `llm`            | object    | Required    | -             | Configuration information for AI service provider  |
 | `apis`           | object    | Required    | -             | Configuration information for external API service provider  |
 | `promptTemplate` | object    | Optional    | -             | Configuration information for Agent ReAct template  |
+| `jsonResp`       | object    | Optional    | -             | Configuring json formatting information  |

 The configuration fields for `llm` are as follows:  
 | Name               | Data Type | Requirement | Default Value | Description                         |
@@ -71,7 +72,13 @@ The configuration fields for `chTemplate` and `enTemplate` are as follows:
 | `observation`   | string    | Optional    | -             | The observation part of the Agent ReAct template     |
 | `thought2`      | string    | Optional    | -             | The thought2 part of the Agent ReAct template       |

-## Usage Example
+The configuration fields for `jsonResp` are as follows:  
+| Name               | Data Type | Requirement | Default Value | Description                         |
+|--------------------|-----------|-------------|---------------|------------------------------------|
+| `enable`           | bool      | Optional    | -             | Whether to enable json formatting.  |
+| `jsonSchema`       | string    | Optional    | -             | Custom json schema               |
+
+## Usage Example-disable json formatting
 **Configuration Information**  
 ```yaml  
 llm:  
@@ -335,3 +342,68 @@ curl 'http://<replace with gateway public IP>/api/openai/v1/chat/completions' \
 {"id":"65dcf12c-61ff-9e68-bffa-44fc9e6070d5","choices":[{"index":0,"message":{"role":"assistant","content":" The German translation of \"Hail Hydra!\" is \"Hoch lebe Hydra!\"."},"finish_reason":"stop"}],"created":1724043865,"model":"qwen-max-0403","object":"chat.completion","usage":{"prompt_tokens":908,"completion_tokens":52,"total_tokens":960}}  
 ```  

+## Usage Example-enable json formatting
+**Configuration Information**  
+Add jsonResp configuration to the above configuration
+```yaml
+jsonResp:
+  enable: true
+```
+
+**Request Example**  
+```shell
+curl 'http://<replace with gateway public IP>/api/openai/v1/chat/completions' \
+-H 'Accept: application/json, text/event-stream' \
+-H 'Content-Type: application/json' \
+--data-raw '{"model":"qwen","frequency_penalty":0,"max_tokens":800,"stream":false,"messages":[{"role":"user","content":"What is the current weather in Beijing ?"}],"presence_penalty":0,"temperature":0,"top_p":0}'
+```
+
+**Response Example**
+
+```json
+{"id":"ebd6ea91-8e38-9e14-9a5b-90178d2edea4","choices":[{"index":0,"message":{"role":"assistant","content": "{\"city\": \"BeiJing\", \"weather_condition\": \"cloudy\", \"temperature\": \"19℃\", \"data_update_time\": \"Oct 9, 2024, at 16:37\"}"},"finish_reason":"stop"}],"created":1723187991,"model":"qwen-max-0403","object":"chat.completion","usage":{"prompt_tokens":890,"completion_tokens":56,"total_tokens":946}}
+```
+If you don't customise the json schema, the big model will automatically generate a json format
+
+**Configuration Information**
+Add custom json schema configuration
+```yaml
+jsonResp:
+  enable: true
+  jsonSchema:
+    title: WeatherSchema
+    type: object
+    properties:
+      location:
+        type: string
+        description: city name.
+      weather:
+        type: string
+        description: weather conditions.
+      temperature:
+        type: string
+        description: temperature.
+      update_time:
+        type: string
+        description: the update time of data.
+    required:
+      - location
+      - weather
+      - temperature
+    additionalProperties: false
+```
+
+**Request Example**
+
+```shell
+curl 'http://<replace with gateway public IP>/api/openai/v1/chat/completions' \
+-H 'Accept: application/json, text/event-stream' \
+-H 'Content-Type: application/json' \
+--data-raw '{"model":"qwen","frequency_penalty":0,"max_tokens":800,"stream":false,"messages":[{"role":"user","content":"What is the current weather in Beijing ?"}],"presence_penalty":0,"temperature":0,"top_p":0}'
+```
+
+**Response Example**
+
+```json
+{"id":"ebd6ea91-8e38-9e14-9a5b-90178d2edea4","choices":[{"index":0,"message":{"role":"assistant","content": "{\"location\": \"Beijing\", \"weather\": \"cloudy\", \"temperature\": \"19℃\", \"update_time\": \"Oct 9, 2024, at 16:37\"}"},"finish_reason":"stop"}],"created":1723187991,"model":"qwen-max-0403","object":"chat.completion","usage":{"prompt_tokens":890,"completion_tokens":56,"total_tokens":946}}
+```
--- a/plugins/wasm-go/extensions/ai-agent/config.go
+++ b/plugins/wasm-go/extensions/ai-agent/config.go
@@ -211,6 +211,15 @@ type LLMInfo struct {
 	MaxTokens int64 `yaml:"maxToken" json:"maxTokens"`
 }

+type JsonResp struct {
+	// @Title zh-CN Enable
+	// @Description zh-CN 是否要启用json格式化输出
+	Enable bool `yaml:"enable" json:"enable"`
+	// @Title zh-CN Json Schema
+	// @Description zh-CN 用以验证响应json的Json Schema, 为空则只验证返回的响应是否为合法json
+	JsonSchema map[string]interface{} `required:"false" json:"jsonSchema" yaml:"jsonSchema"`
+}
+
 type PluginConfig struct {
 	// @Title zh-CN 返回 HTTP 响应的模版
 	// @Description zh-CN 用 %s 标记需要被 cache value 替换的部分
@@ -225,6 +234,7 @@ type PluginConfig struct {
 	LLMClient      wrapper.HttpClient `yaml:"-" json:"-"`
 	APIsParam      []APIsParam        `yaml:"-" json:"-"`
 	PromptTemplate PromptTemplate     `yaml:"promptTemplate" json:"promptTemplate"`
+	JsonResp       JsonResp           `yaml:"jsonResp" json:"jsonResp"`
 }

 func initResponsePromptTpl(gjson gjson.Result, c *PluginConfig) {
@@ -402,3 +412,15 @@ func initLLMClient(gjson gjson.Result, c *PluginConfig) {
 		Host: c.LLMInfo.Domain,
 	})
 }
+
+func initJsonResp(gjson gjson.Result, c *PluginConfig) {
+	c.JsonResp.Enable = false
+	if c.JsonResp.Enable = gjson.Get("jsonResp.enable").Bool(); c.JsonResp.Enable {
+		c.JsonResp.JsonSchema = nil
+		if jsonSchemaValue := gjson.Get("jsonResp.jsonSchema"); jsonSchemaValue.Exists() {
+			if schemaValue, ok := jsonSchemaValue.Value().(map[string]interface{}); ok {
+				c.JsonResp.JsonSchema = schemaValue
+			}
+		}
+	}
+}
--- a/plugins/wasm-go/extensions/ai-agent/main.go
+++ b/plugins/wasm-go/extensions/ai-agent/main.go
@@ -2,8 +2,10 @@ package main

 import (
 	"encoding/json"
+	"errors"
 	"fmt"
 	"net/http"
+	"net/url"
 	"regexp"
 	"strings"

@@ -47,6 +49,8 @@ func parseConfig(gjson gjson.Result, c *PluginConfig, log wrapper.Log) error {

 	initLLMClient(gjson, c)

+	initJsonResp(gjson, c)
+
 	return nil
 }

@@ -76,10 +80,10 @@ func firstReq(ctx wrapper.HttpContext, config PluginConfig, prompt string, rawRe
 		log.Debugf("[onHttpRequestBody] newRequestBody: %s", string(newbody))
 		err := proxywasm.ReplaceHttpRequestBody(newbody)
 		if err != nil {
-			log.Debug("替换失败")
+			log.Debugf("failed replace err: %s", err.Error())
 			proxywasm.SendHttpResponse(200, [][2]string{{"content-type", "application/json; charset=utf-8"}}, []byte(fmt.Sprintf(config.ReturnResponseTemplate, "替换失败"+err.Error())), -1)
 		}
-		log.Debug("[onHttpRequestBody] request替换成功")
+		log.Debug("[onHttpRequestBody] replace request success")
 		return types.ActionContinue
 	}
 }
@@ -175,11 +179,103 @@ func onHttpResponseHeaders(ctx wrapper.HttpContext, config PluginConfig, log wra
 	return types.ActionContinue
 }

-func toolsCallResult(ctx wrapper.HttpContext, config PluginConfig, content string, rawResponse Response, log wrapper.Log, statusCode int, responseBody []byte) {
+func extractJson(bodyStr string) (string, error) {
+	// simply extract json from response body string
+	startIndex := strings.Index(bodyStr, "{")
+	endIndex := strings.LastIndex(bodyStr, "}") + 1
+
+	// if not found
+	if startIndex == -1 || startIndex >= endIndex {
+		return "", errors.New("cannot find json in the response body")
+	}
+
+	jsonStr := bodyStr[startIndex:endIndex]
+
+	// attempt to parse the JSON
+	var result map[string]interface{}
+	err := json.Unmarshal([]byte(jsonStr), &result)
+	if err != nil {
+		return "", err
+	}
+	return jsonStr, nil
+}
+
+func jsonFormat(llmClient wrapper.HttpClient, llmInfo LLMInfo, jsonSchema map[string]interface{}, assistantMessage Message, actionInput string, headers [][2]string, streamMode bool, rawResponse Response, log wrapper.Log) string {
+	prompt := fmt.Sprintf(prompttpl.Json_Resp_Template, jsonSchema, actionInput)
+
+	messages := []dashscope.Message{{Role: "user", Content: prompt}}
+
+	completion := dashscope.Completion{
+		Model:    llmInfo.Model,
+		Messages: messages,
+	}
+
+	completionSerialized, _ := json.Marshal(completion)
+	var content string
+	err := llmClient.Post(
+		llmInfo.Path,
+		headers,
+		completionSerialized,
+		func(statusCode int, responseHeaders http.Header, responseBody []byte) {
+			//得到gpt的返回结果
+			var responseCompletion dashscope.CompletionResponse
+			_ = json.Unmarshal(responseBody, &responseCompletion)
+			log.Infof("[jsonFormat] content: %s", responseCompletion.Choices[0].Message.Content)
+			content = responseCompletion.Choices[0].Message.Content
+			jsonStr, err := extractJson(content)
+			if err != nil {
+				log.Debugf("[onHttpRequestBody] extractJson err: %s", err.Error())
+				jsonStr = content
+			}
+
+			if streamMode {
+				stream(jsonStr, rawResponse, log)
+			} else {
+				noneStream(assistantMessage, jsonStr, rawResponse, log)
+			}
+		}, uint32(llmInfo.MaxExecutionTime))
+	if err != nil {
+		log.Debugf("[onHttpRequestBody] completion err: %s", err.Error())
+		proxywasm.ResumeHttpRequest()
+	}
+	return content
+}
+
+func noneStream(assistantMessage Message, actionInput string, rawResponse Response, log wrapper.Log) {
+	assistantMessage.Role = "assistant"
+	assistantMessage.Content = actionInput
+	rawResponse.Choices[0].Message = assistantMessage
+	newbody, err := json.Marshal(rawResponse)
+	if err != nil {
+		proxywasm.ResumeHttpResponse()
+		return
+	} else {
+		proxywasm.ReplaceHttpResponseBody(newbody)
+
+		log.Debug("[onHttpResponseBody] replace response success")
+		proxywasm.ResumeHttpResponse()
+	}
+}
+
+func stream(actionInput string, rawResponse Response, log wrapper.Log) {
+	headers := [][2]string{{"content-type", "text/event-stream; charset=utf-8"}}
+	proxywasm.ReplaceHttpResponseHeaders(headers)
+	// Remove quotes from actionInput
+	actionInput = strings.Trim(actionInput, "\"")
+	returnStreamResponseTemplate := `data:{"id":"%s","choices":[{"index":0,"delta":{"role":"assistant","content":"%s"},"finish_reason":"stop"}],"model":"%s","object":"chat.completion","usage":{"prompt_tokens":%d,"completion_tokens":%d,"total_tokens":%d}}` + "\n\ndata:[DONE]\n\n"
+	newbody := fmt.Sprintf(returnStreamResponseTemplate, rawResponse.ID, actionInput, rawResponse.Model, rawResponse.Usage.PromptTokens, rawResponse.Usage.CompletionTokens, rawResponse.Usage.TotalTokens)
+	log.Infof("[onHttpResponseBody] newResponseBody: ", newbody)
+	proxywasm.ReplaceHttpResponseBody([]byte(newbody))
+
+	log.Debug("[onHttpResponseBody] replace response success")
+	proxywasm.ResumeHttpResponse()
+}
+
+func toolsCallResult(ctx wrapper.HttpContext, llmClient wrapper.HttpClient, llmInfo LLMInfo, jsonResp JsonResp, aPIsParam []APIsParam, aPIClient []wrapper.HttpClient, content string, rawResponse Response, log wrapper.Log, statusCode int, responseBody []byte) {
 	if statusCode != http.StatusOK {
 		log.Debugf("statusCode: %d", statusCode)
 	}
-	log.Info("========函数返回结果========")
+	log.Info("========function result========")
 	log.Infof(string(responseBody))

 	observation := "Observation: " + string(responseBody)
@@ -187,15 +283,15 @@ func toolsCallResult(ctx wrapper.HttpContext, config PluginConfig, content strin
 	dashscope.MessageStore.AddForUser(observation)

 	completion := dashscope.Completion{
-		Model:     config.LLMInfo.Model,
+		Model:     llmInfo.Model,
 		Messages:  dashscope.MessageStore,
-		MaxTokens: config.LLMInfo.MaxTokens,
+		MaxTokens: llmInfo.MaxTokens,
 	}

-	headers := [][2]string{{"Content-Type", "application/json"}, {"Authorization", "Bearer " + config.LLMInfo.APIKey}}
+	headers := [][2]string{{"Content-Type", "application/json"}, {"Authorization", "Bearer " + llmInfo.APIKey}}
 	completionSerialized, _ := json.Marshal(completion)
-	err := config.LLMClient.Post(
-		config.LLMInfo.Path,
+	err := llmClient.Post(
+		llmInfo.Path,
 		headers,
 		completionSerialized,
 		func(statusCode int, responseHeaders http.Header, responseBody []byte) {
@@ -205,42 +301,31 @@ func toolsCallResult(ctx wrapper.HttpContext, config PluginConfig, content strin
 			log.Infof("[toolsCall] content: %s", responseCompletion.Choices[0].Message.Content)

 			if responseCompletion.Choices[0].Message.Content != "" {
-				retType, actionInput := toolsCall(ctx, config, responseCompletion.Choices[0].Message.Content, rawResponse, log)
+				retType, actionInput := toolsCall(ctx, llmClient, llmInfo, jsonResp, aPIsParam, aPIClient, responseCompletion.Choices[0].Message.Content, rawResponse, log)
 				if retType == types.ActionContinue {
 					//得到了Final Answer
 					var assistantMessage Message
+					var streamMode bool
 					if ctx.GetContext(StreamContextKey) == nil {
-						assistantMessage.Role = "assistant"
-						assistantMessage.Content = actionInput
-						rawResponse.Choices[0].Message = assistantMessage
-						newbody, err := json.Marshal(rawResponse)
-						if err != nil {
-							proxywasm.ResumeHttpResponse()
-							return
+						streamMode = false
+						if jsonResp.Enable {
+							jsonFormat(llmClient, llmInfo, jsonResp.JsonSchema, assistantMessage, actionInput, headers, streamMode, rawResponse, log)
 						} else {
-							proxywasm.ReplaceHttpResponseBody(newbody)
-
-							log.Debug("[onHttpResponseBody] response替换成功")
-							proxywasm.ResumeHttpResponse()
+							noneStream(assistantMessage, actionInput, rawResponse, log)
 						}
 					} else {
-						headers := [][2]string{{"content-type", "text/event-stream; charset=utf-8"}}
-						proxywasm.ReplaceHttpResponseHeaders(headers)
-						// Remove quotes from actionInput
-						actionInput = strings.Trim(actionInput, "\"")
-						returnStreamResponseTemplate := `data:{"id":"%s","choices":[{"index":0,"delta":{"role":"assistant","content":"%s"},"finish_reason":"stop"}],"model":"%s","object":"chat.completion","usage":{"prompt_tokens":%d,"completion_tokens":%d,"total_tokens":%d}}` + "\n\ndata:[DONE]\n\n"
-						newbody := fmt.Sprintf(returnStreamResponseTemplate, rawResponse.ID, actionInput, rawResponse.Model, rawResponse.Usage.PromptTokens, rawResponse.Usage.CompletionTokens, rawResponse.Usage.TotalTokens)
-						log.Infof("[onHttpResponseBody] newResponseBody: ", newbody)
-						proxywasm.ReplaceHttpResponseBody([]byte(newbody))
-
-						log.Debug("[onHttpResponseBody] response替换成功")
-						proxywasm.ResumeHttpResponse()
+						streamMode = true
+						if jsonResp.Enable {
+							jsonFormat(llmClient, llmInfo, jsonResp.JsonSchema, assistantMessage, actionInput, headers, streamMode, rawResponse, log)
+						} else {
+							stream(actionInput, rawResponse, log)
+						}
 					}
 				}
 			} else {
 				proxywasm.ResumeHttpRequest()
 			}
-		}, uint32(config.LLMInfo.MaxExecutionTime))
+		}, uint32(llmInfo.MaxExecutionTime))
 	if err != nil {
 		log.Debugf("[onHttpRequestBody] completion err: %s", err.Error())
 		proxywasm.ResumeHttpRequest()
@@ -294,7 +379,7 @@ func outputParser(response string, log wrapper.Log) (string, string) {
 	return "", ""
 }

-func toolsCall(ctx wrapper.HttpContext, config PluginConfig, content string, rawResponse Response, log wrapper.Log) (types.Action, string) {
+func toolsCall(ctx wrapper.HttpContext, llmClient wrapper.HttpClient, llmInfo LLMInfo, jsonResp JsonResp, aPIsParam []APIsParam, aPIClient []wrapper.HttpClient, content string, rawResponse Response, log wrapper.Log) (types.Action, string) {
 	dashscope.MessageStore.AddForAssistant(content)

 	action, actionInput := outputParser(content, log)
@@ -305,9 +390,9 @@ func toolsCall(ctx wrapper.HttpContext, config PluginConfig, content string, raw
 	}
 	count := ctx.GetContext(ToolCallsCount).(int)
 	count++
-	log.Debugf("toolCallsCount:%d, config.LLMInfo.MaxIterations=%d", count, config.LLMInfo.MaxIterations)
+	log.Debugf("toolCallsCount:%d, config.LLMInfo.MaxIterations=%d", count, llmInfo.MaxIterations)
 	//函数递归调用次数，达到了预设的循环次数，强制结束
-	if int64(count) > config.LLMInfo.MaxIterations {
+	if int64(count) > llmInfo.MaxIterations {
 		ctx.SetContext(ToolCallsCount, 0)
 		return types.ActionContinue, ""
 	} else {
@@ -316,15 +401,14 @@ func toolsCall(ctx wrapper.HttpContext, config PluginConfig, content string, raw

 	//没得到最终答案

-	var url string
+	var urlStr string
 	var headers [][2]string
 	var apiClient wrapper.HttpClient
 	var method string
 	var reqBody []byte
-	var key string
 	var maxExecutionTime int64

-	for i, apisParam := range config.APIsParam {
+	for i, apisParam := range aPIsParam {
 		maxExecutionTime = apisParam.MaxExecutionTime
 		for _, tools_param := range apisParam.ToolsParam {
 			if action == tools_param.ToolName {
@@ -340,28 +424,37 @@ func toolsCall(ctx wrapper.HttpContext, config PluginConfig, content string, raw

 				method = tools_param.Method

-				// 组装 headers 和 key
-				headers = [][2]string{{"Content-Type", "application/json"}}
-				if apisParam.APIKey.Name != "" {
-					if apisParam.APIKey.In == "query" {
-						key = "?" + apisParam.APIKey.Name + "=" + apisParam.APIKey.Value
-					} else if apisParam.APIKey.In == "header" {
-						headers = append(headers, [2]string{"Authorization", apisParam.APIKey.Name + " " + apisParam.APIKey.Value})
+				// 组装 URL 和请求体
+				urlStr = apisParam.URL + tools_param.Path
+
+				// 解析URL模板以查找路径参数
+				urlParts := strings.Split(urlStr, "/")
+				for i, part := range urlParts {
+					if strings.Contains(part, "{") && strings.Contains(part, "}") {
+						for _, param := range tools_param.ParamName {
+							paramNameInPath := part[1 : len(part)-1]
+							if paramNameInPath == param {
+								if value, ok := data[param]; ok {
+									// 删除已经使用过的
+									delete(data, param)
+									// 替换模板中的占位符
+									urlParts[i] = url.QueryEscape(value.(string))
+								}
+							}
+						}
 					}
 				}

-				// 组装 URL 和请求体
-				url = apisParam.URL + tools_param.Path + key
+				// 重新组合URL
+				urlStr = strings.Join(urlParts, "/")
+
+				queryParams := make([][2]string, 0)
 				if method == "GET" {
-					queryParams := make([]string, 0, len(tools_param.ParamName))
 					for _, param := range tools_param.ParamName {
 						if value, ok := data[param]; ok {
-							queryParams = append(queryParams, fmt.Sprintf("%s=%v", param, value))
+							queryParams = append(queryParams, [2]string{param, fmt.Sprintf("%v", value)})
 						}
 					}
-					if len(queryParams) > 0 {
-						url += "&" + strings.Join(queryParams, "&")
-					}
 				} else if method == "POST" {
 					var err error
 					reqBody, err = json.Marshal(data)
@@ -371,9 +464,30 @@ func toolsCall(ctx wrapper.HttpContext, config PluginConfig, content string, raw
 					}
 				}

-				log.Infof("url: %s", url)
+				// 组装 headers 和 key
+				headers = [][2]string{{"Content-Type", "application/json"}}
+				if apisParam.APIKey.Name != "" {
+					if apisParam.APIKey.In == "query" {
+						queryParams = append(queryParams, [2]string{apisParam.APIKey.Name, apisParam.APIKey.Value})
+					} else if apisParam.APIKey.In == "header" {
+						headers = append(headers, [2]string{"Authorization", apisParam.APIKey.Name + " " + apisParam.APIKey.Value})
+					}
+				}

-				apiClient = config.APIClient[i]
+				if len(queryParams) > 0 {
+					// 将 key 拼接到 url 后面
+					urlStr += "?"
+					for i, param := range queryParams {
+						if i != 0 {
+							urlStr += "&"
+						}
+						urlStr += url.QueryEscape(param[0]) + "=" + url.QueryEscape(param[1])
+					}
+				}
+
+				log.Debugf("url: %s", urlStr)
+
+				apiClient = aPIClient[i]
 				break
 			}
 		}
@@ -382,11 +496,11 @@ func toolsCall(ctx wrapper.HttpContext, config PluginConfig, content string, raw
 	if apiClient != nil {
 		err := apiClient.Call(
 			method,
-			url,
+			urlStr,
 			headers,
 			reqBody,
 			func(statusCode int, responseHeaders http.Header, responseBody []byte) {
-				toolsCallResult(ctx, config, content, rawResponse, log, statusCode, responseBody)
+				toolsCallResult(ctx, llmClient, llmInfo, jsonResp, aPIsParam, aPIClient, content, rawResponse, log, statusCode, responseBody)
 			}, uint32(maxExecutionTime))
 		if err != nil {
 			log.Debugf("tool calls error: %s", err.Error())
@@ -415,7 +529,7 @@ func onHttpResponseBody(ctx wrapper.HttpContext, config PluginConfig, body []byt
 	//如果gpt返回的内容不是空的
 	if rawResponse.Choices[0].Message.Content != "" {
 		//进入agent的循环思考，工具调用的过程中
-		retType, _ := toolsCall(ctx, config, rawResponse.Choices[0].Message.Content, rawResponse, log)
+		retType, _ := toolsCall(ctx, config.LLMClient, config.LLMInfo, config.JsonResp, config.APIsParam, config.APIClient, rawResponse.Choices[0].Message.Content, rawResponse, log)
 		return retType
 	} else {
 		return types.ActionContinue
--- a/plugins/wasm-go/extensions/ai-agent/promptTpl/prompt.go
+++ b/plugins/wasm-go/extensions/ai-agent/promptTpl/prompt.go
@@ -167,3 +167,7 @@ Action:` + "```" + `
 %s
 Question: %s
 `
+const Json_Resp_Template = `
+Given the Json Schema: %s, please help me convert the following content to a pure json: %s
+Do not respond other content except the pure json!!!!
+`
--- a/plugins/wasm-go/extensions/ai-cache/README.md
+++ b/plugins/wasm-go/extensions/ai-cache/README.md
@@ -60,7 +60,7 @@ LLM 结果缓存插件，默认配置方式可以直接用于 openai 协议的
 | vector.apiKey | string | optional | ""  | 向量存储服务 API Key |
 | vector.topK | int | optional | 1 | 返回TopK结果，默认为 1 |
 | vector.timeout | uint32 | optional | 10000 | 请求向量存储服务的超时时间，单位为毫秒。默认值是10000，即10秒 |
-| vector.collectionID | string | optional | "" |  dashvector 向量存储服务 Collection ID |
+| vector.collectionID | string | optional | "" | 向量存储服务 Collection ID |
 | vector.threshold | float64 | optional | 1000 | 向量相似度度量阈值 |
 | vector.thresholdRelation | string | optional | lt | 相似度度量方式有 `Cosine`, `DotProduct`, `Euclidean` 等，前两者值越大相似度越高，后者值越小相似度越高。对于 `Cosine` 和 `DotProduct` 选择 `gt`，对于 `Euclidean` 则选择 `lt`。默认为 `lt`，所有条件包括 `lt` (less than，小于)、`lte` (less than or equal to，小等于)、`gt` (greater than，大于)、`gte` (greater than or equal to，大等于) |

@@ -99,6 +99,45 @@ LLM 结果缓存插件，默认配置方式可以直接用于 openai 协议的
 | responseTemplate | string | optional | `{"id":"ai-cache.hit","choices":[{"index":0,"message":{"role":"assistant","content":%s},"finish_reason":"stop"}],"model":"gpt-4o","object":"chat.completion","usage":{"prompt_tokens":0,"completion_tokens":0,"total_tokens":0}}` | 返回 HTTP 响应的模版，用 %s 标记需要被 cache value 替换的部分 |
 | streamResponseTemplate | string | optional | `data:{"id":"ai-cache.hit","choices":[{"index":0,"delta":{"role":"assistant","content":%s},"finish_reason":"stop"}],"model":"gpt-4o","object":"chat.completion","usage":{"prompt_tokens":0,"completion_tokens":0,"total_tokens":0}}\n\ndata:[DONE]\n\n` | 返回流式 HTTP 响应的模版，用 %s 标记需要被 cache value 替换的部分 |

+# 向量数据库提供商特有配置
+## Chroma
+Chroma 所对应的 `vector.type` 为 `chroma`。它并无特有的配置字段。需要提前创建 Collection，并填写 Collection ID 至配置项 `vector.collectionID`，一个 Collection ID 的示例为 `52bbb8b3-724c-477b-a4ce-d5b578214612`。
+
+## DashVector
+DashVector 所对应的 `vector.type` 为 `dashvector`。它并无特有的配置字段。需要提前创建 Collection，并填写 `Collection 名称` 至配置项 `vector.collectionID`。
+
+## ElasticSearch
+ElasticSearch 所对应的 `vector.type` 为 `elasticsearch`。需要提前创建 Index 并填写 Index Name 至配置项 `vector.collectionID` 。
+
+当前依赖于 [KNN](https://www.elastic.co/guide/en/elasticsearch/reference/current/knn-search.html) 方法，请保证 ES 版本支持 `KNN`，当前已在 `8.16` 版本测试。
+
+它特有的配置字段如下：
+| 名称              | 数据类型 | 填写要求 | 默认值 | 描述                                                                          |
+|-------------------|----------|----------|--------|-------------------------------------------------------------------------------|
+| `vector.esUsername` | string   | 非必填   | -      | ElasticSearch 用户名 |
+| `vector.esPassword` | string | 非必填 | - | ElasticSearch 密码 |
+
+
+`vector.esUsername` 和 `vector.esPassword` 用于 Basic 认证。同时也支持 Api Key 认证，当填写了 `vector.apiKey` 时，则启用 Api Key 认证，如果使用 SaaS 版本需要填写 `encoded` 的值。
+
+## Milvus
+Milvus 所对应的 `vector.type` 为 `milvus`。它并无特有的配置字段。需要提前创建 Collection，并填写 Collection Name 至配置项 `vector.collectionID`。
+
+## Pinecone
+Pinecone 所对应的 `vector.type` 为 `pinecone`。它并无特有的配置字段。需要提前创建 Index，并填写 Index 访问域名至 `vector.serviceHost`。
+
+Pinecone 中的 `Namespace` 参数通过插件的 `vector.collectionID` 进行配置，如果不填写 `vector.collectionID`，则默认为 Default Namespace。
+
+## Qdrant
+Qdrant 所对应的 `vector.type` 为 `qdrant`。它并无特有的配置字段。需要提前创建 Collection，并填写 Collection Name 至配置项 `vector.collectionID`。
+
+## Weaviate
+Weaviate 所对应的 `vector.type` 为 `weaviate`。它并无特有的配置字段。
+需要提前创建 Collection，并填写 Collection Name 至配置项 `vector.collectionID`。
+
+需要注意的是 Weaviate 会设置首字母自动大写，在填写配置 `collectionID` 的时候需要将首字母设置为大写。
+
+如果使用 SaaS 需要填写 `vector.serviceHost` 参数。

 ## 配置示例
 ### 基础配置
@@ -144,4 +183,4 @@ GJSON PATH 支持条件判断语法，例如希望取最后一个 role 为 user

 ## 常见问题

-1. 如果返回的错误为 `error status returned by host: bad argument`，请检查`serviceName`是否正确包含了服务的类型后缀(.dns等)。
+1. 如果返回的错误为 `error status returned by host: bad argument`，请检查`serviceName`是否正确包含了服务的类型后缀(.dns等)。
--- a/plugins/wasm-go/extensions/ai-cache/cache/provider.go
+++ b/plugins/wasm-go/extensions/ai-cache/cache/provider.go
@@ -2,6 +2,7 @@ package cache

 import (
 	"errors"
+	"strings"

 	"github.com/alibaba/higress/plugins/wasm-go/pkg/wrapper"
 	"github.com/tidwall/gjson"
@@ -62,7 +63,12 @@ func (c *ProviderConfig) FromJson(json gjson.Result) {
 	c.serviceName = json.Get("serviceName").String()
 	c.servicePort = int(json.Get("servicePort").Int())
 	if !json.Get("servicePort").Exists() {
-		c.servicePort = 6379
+		if strings.HasSuffix(c.serviceName, ".static") {
+			// use default logic port which is 80 for static service
+			c.servicePort = 80
+		} else {
+			c.servicePort = 6379
+		}
 	}
 	c.serviceHost = json.Get("serviceHost").String()
 	c.username = json.Get("username").String()
--- a/plugins/wasm-go/extensions/ai-cache/config/config.go
+++ b/plugins/wasm-go/extensions/ai-cache/config/config.go
@@ -79,11 +79,11 @@ func (c *PluginConfig) FromJson(json gjson.Result, log wrapper.Log) {

 	c.StreamResponseTemplate = json.Get("streamResponseTemplate").String()
 	if c.StreamResponseTemplate == "" {
-		c.StreamResponseTemplate = `data:{"id":"from-cache","choices":[{"index":0,"delta":{"role":"assistant","content":"%s"},"finish_reason":"stop"}],"model":"gpt-4o","object":"chat.completion","usage":{"prompt_tokens":0,"completion_tokens":0,"total_tokens":0}}` + "\n\ndata:[DONE]\n\n"
+		c.StreamResponseTemplate = `data:{"id":"from-cache","choices":[{"index":0,"delta":{"role":"assistant","content":"%s"},"finish_reason":"stop"}],"model":"from-cache","object":"chat.completion","usage":{"prompt_tokens":0,"completion_tokens":0,"total_tokens":0}}` + "\n\ndata:[DONE]\n\n"
 	}
 	c.ResponseTemplate = json.Get("responseTemplate").String()
 	if c.ResponseTemplate == "" {
-		c.ResponseTemplate = `{"id":"from-cache","choices":[{"index":0,"message":{"role":"assistant","content":"%s"},"finish_reason":"stop"}],"model":"gpt-4o","object":"chat.completion","usage":{"prompt_tokens":0,"completion_tokens":0,"total_tokens":0}}`
+		c.ResponseTemplate = `{"id":"from-cache","choices":[{"index":0,"message":{"role":"assistant","content":"%s"},"finish_reason":"stop"}],"model":"from-cache","object":"chat.completion","usage":{"prompt_tokens":0,"completion_tokens":0,"total_tokens":0}}`
 	}

 	if json.Get("enableSemanticCache").Exists() {
--- a/plugins/wasm-go/extensions/ai-cache/core.go
+++ b/plugins/wasm-go/extensions/ai-cache/core.go
@@ -74,6 +74,9 @@ func processCacheHit(key string, response string, stream bool, ctx wrapper.HttpC

 	ctx.SetContext(CACHE_KEY_CONTEXT_KEY, nil)

+	ctx.SetUserAttribute("cache_status", "hit")
+	ctx.WriteUserAttributeToLogWithKey(wrapper.AILogKey)
+
 	if stream {
 		proxywasm.SendHttpResponseWithDetail(200, "ai-cache.hit", [][2]string{{"content-type", "text/event-stream; charset=utf-8"}}, []byte(fmt.Sprintf(c.StreamResponseTemplate, escapedResponse)), -1)
 	} else {
--- a/plugins/wasm-go/extensions/ai-cache/embedding/cohere.go
+++ b/plugins/wasm-go/extensions/ai-cache/embedding/cohere.go
@@ -0,0 +1,158 @@
+package embedding
+
+import (
+	"encoding/json"
+	"errors"
+	"fmt"
+	"net/http"
+	"strconv"
+
+	"github.com/alibaba/higress/plugins/wasm-go/pkg/wrapper"
+	"github.com/tidwall/gjson"
+)
+
+const (
+	COHERE_DOMAIN             = "api.cohere.com"
+	COHERE_PORT               = 443
+	COHERE_DEFAULT_MODEL_NAME = "embed-english-v2.0"
+	COHERE_ENDPOINT           = "/v2/embed"
+)
+
+type cohereProviderInitializer struct {
+}
+
+var cohereConfig cohereProviderConfig
+
+type cohereProviderConfig struct {
+	// @Title zh-CN 文本特征提取服务 API Key
+	// @Description zh-CN 文本特征提取服务 API Key
+	apiKey string
+}
+
+func (c *cohereProviderInitializer) InitConfig(json gjson.Result) {
+	cohereConfig.apiKey = json.Get("apiKey").String()
+}
+func (c *cohereProviderInitializer) ValidateConfig() error {
+	if cohereConfig.apiKey == "" {
+		return errors.New("[Cohere] apiKey is required")
+	}
+	return nil
+}
+
+func (t *cohereProviderInitializer) CreateProvider(c ProviderConfig) (Provider, error) {
+	if c.servicePort == 0 {
+		c.servicePort = COHERE_PORT
+	}
+	if c.serviceHost == "" {
+		c.serviceHost = COHERE_DOMAIN
+	}
+	return &CohereProvider{
+		config: c,
+		client: wrapper.NewClusterClient(wrapper.FQDNCluster{
+			FQDN: c.serviceName,
+			Host: c.serviceHost,
+			Port: int64(c.servicePort),
+		}),
+	}, nil
+}
+
+type cohereResponse struct {
+	Embeddings cohereEmbeddings `json:"embeddings"`
+}
+
+type cohereEmbeddings struct {
+	FloatTypeEebedding [][]float64 `json:"float"`
+}
+
+type cohereEmbeddingRequest struct {
+	Texts          []string `json:"texts"`
+	Model          string   `json:"model"`
+	InputType      string   `json:"input_type"`
+	EmbeddingTypes []string `json:"embedding_types"`
+}
+
+type CohereProvider struct {
+	config ProviderConfig
+	client wrapper.HttpClient
+}
+
+func (t *CohereProvider) GetProviderType() string {
+	return PROVIDER_TYPE_COHERE
+}
+func (t *CohereProvider) constructParameters(texts []string, log wrapper.Log) (string, [][2]string, []byte, error) {
+	model := t.config.model
+
+	if model == "" {
+		model = COHERE_DEFAULT_MODEL_NAME
+	}
+	data := cohereEmbeddingRequest{
+		Texts:          texts,
+		Model:          model,
+		InputType:      "search_document",
+		EmbeddingTypes: []string{"float"},
+	}
+
+	requestBody, err := json.Marshal(data)
+	if err != nil {
+		log.Errorf("failed to marshal request data: %v", err)
+		return "", nil, nil, err
+	}
+
+	headers := [][2]string{
+		{"Authorization", fmt.Sprintf("BEARER %s", cohereConfig.apiKey)},
+		{"Content-Type", "application/json"},
+	}
+
+	return COHERE_ENDPOINT, headers, requestBody, nil
+}
+
+func (t *CohereProvider) parseTextEmbedding(responseBody []byte) (*cohereResponse, error) {
+	var resp cohereResponse
+	err := json.Unmarshal(responseBody, &resp)
+	if err != nil {
+		return nil, err
+	}
+	return &resp, nil
+}
+
+func (t *CohereProvider) GetEmbedding(
+	queryString string,
+	ctx wrapper.HttpContext,
+	log wrapper.Log,
+	callback func(emb []float64, err error)) error {
+	embUrl, embHeaders, embRequestBody, err := t.constructParameters([]string{queryString}, log)
+	if err != nil {
+		log.Errorf("failed to construct parameters: %v", err)
+		return err
+	}
+
+	var resp *cohereResponse
+	err = t.client.Post(embUrl, embHeaders, embRequestBody,
+		func(statusCode int, responseHeaders http.Header, responseBody []byte) {
+
+			if statusCode != http.StatusOK {
+				err = errors.New("failed to get embedding due to status code: " + strconv.Itoa(statusCode))
+				callback(nil, err)
+				return
+			}
+
+			log.Debugf("get embedding response: %d, %s", statusCode, responseBody)
+
+			resp, err = t.parseTextEmbedding(responseBody)
+			if err != nil {
+				err = fmt.Errorf("failed to parse response: %v", err)
+				callback(nil, err)
+				return
+			}
+
+			if len(resp.Embeddings.FloatTypeEebedding) == 0 {
+				err = errors.New("no embedding found in response")
+				callback(nil, err)
+				return
+			}
+
+			callback(resp.Embeddings.FloatTypeEebedding[0], nil)
+
+		}, t.config.timeout)
+	return err
+}
--- a/plugins/wasm-go/extensions/ai-cache/embedding/dashscope.go
+++ b/plugins/wasm-go/extensions/ai-cache/embedding/dashscope.go
@@ -8,6 +8,7 @@ import (
 	"strconv"

 	"github.com/alibaba/higress/plugins/wasm-go/pkg/wrapper"
+	"github.com/tidwall/gjson"
 )

 const (
@@ -17,11 +18,22 @@ const (
 	DASHSCOPE_ENDPOINT           = "/api/v1/services/embeddings/text-embedding/text-embedding"
 )

+var dashScopeConfig dashScopeProviderConfig
+
 type dashScopeProviderInitializer struct {
 }
+type dashScopeProviderConfig struct {
+	// @Title zh-CN 文本特征提取服务 API Key
+	// @Description zh-CN 文本特征提取服务 API Key
+	apiKey string
+}

-func (d *dashScopeProviderInitializer) ValidateConfig(config ProviderConfig) error {
-	if config.apiKey == "" {
+func (c *dashScopeProviderInitializer) InitConfig(json gjson.Result) {
+	dashScopeConfig.apiKey = json.Get("apiKey").String()
+}
+
+func (c *dashScopeProviderInitializer) ValidateConfig() error {
+	if dashScopeConfig.apiKey == "" {
 		return errors.New("[DashScope] apiKey is required")
 	}
 	return nil
@@ -114,14 +126,14 @@ func (d *DSProvider) constructParameters(texts []string, log wrapper.Log) (strin
 		return "", nil, nil, err
 	}

-	if d.config.apiKey == "" {
+	if dashScopeConfig.apiKey == "" {
 		err := errors.New("dashScopeKey is empty")
 		log.Errorf("failed to construct headers: %v", err)
 		return "", nil, nil, err
 	}

 	headers := [][2]string{
-		{"Authorization", "Bearer " + d.config.apiKey},
+		{"Authorization", "Bearer " + dashScopeConfig.apiKey},
 		{"Content-Type", "application/json"},
 	}

--- a/plugins/wasm-go/extensions/ai-cache/embedding/openai.go
+++ b/plugins/wasm-go/extensions/ai-cache/embedding/openai.go
@@ -0,0 +1,170 @@
+package embedding
+
+import (
+	"encoding/json"
+	"errors"
+	"fmt"
+	"net/http"
+
+	"github.com/alibaba/higress/plugins/wasm-go/pkg/wrapper"
+	"github.com/tidwall/gjson"
+)
+
+const (
+	OPENAI_DOMAIN             = "api.openai.com"
+	OPENAI_PORT               = 443
+	OPENAI_DEFAULT_MODEL_NAME = "text-embedding-3-small"
+	OPENAI_ENDPOINT           = "/v1/embeddings"
+)
+
+type openAIProviderInitializer struct {
+}
+
+var openAIConfig openAIProviderConfig
+
+type openAIProviderConfig struct {
+	// @Title zh-CN 文本特征提取服务 API Key
+	// @Description zh-CN 文本特征提取服务 API Key
+	apiKey string
+}
+
+func (c *openAIProviderInitializer) InitConfig(json gjson.Result) {
+	openAIConfig.apiKey = json.Get("apiKey").String()
+}
+
+func (c *openAIProviderInitializer) ValidateConfig() error {
+	if openAIConfig.apiKey == "" {
+		return errors.New("[openAI] apiKey is required")
+	}
+	return nil
+}
+
+func (t *openAIProviderInitializer) CreateProvider(c ProviderConfig) (Provider, error) {
+	if c.servicePort == 0 {
+		c.servicePort = OPENAI_PORT
+	}
+	if c.serviceHost == "" {
+		c.serviceHost = OPENAI_DOMAIN
+	}
+	if c.model == "" {
+		c.model = OPENAI_DEFAULT_MODEL_NAME
+	}
+	return &OpenAIProvider{
+		config: c,
+		client: wrapper.NewClusterClient(wrapper.FQDNCluster{
+			FQDN: c.serviceName,
+			Host: c.serviceHost,
+			Port: c.servicePort,
+		}),
+	}, nil
+}
+
+func (t *OpenAIProvider) GetProviderType() string {
+	return PROVIDER_TYPE_OPENAI
+}
+
+type OpenAIResponse struct {
+	Object string         `json:"object"`
+	Data   []OpenAIResult `json:"data"`
+	Model  string         `json:"model"`
+	Error  *OpenAIError   `json:"error"`
+}
+
+type OpenAIResult struct {
+	Object    string    `json:"object"`
+	Embedding []float64 `json:"embedding"`
+	Index     int       `json:"index"`
+}
+
+type OpenAIError struct {
+	Message string `json:"prompt_tokens"`
+	Type    string `json:"type"`
+	Code    string `json:"code"`
+	Param   string `json:"param"`
+}
+
+type OpenAIEmbeddingRequest struct {
+	Input string `json:"input"`
+	Model string `json:"model"`
+}
+
+type OpenAIProvider struct {
+	config ProviderConfig
+	client wrapper.HttpClient
+}
+
+func (t *OpenAIProvider) constructParameters(text string, log wrapper.Log) (string, [][2]string, []byte, error) {
+	if text == "" {
+		err := errors.New("queryString text cannot be empty")
+		return "", nil, nil, err
+	}
+
+	data := OpenAIEmbeddingRequest{
+		Input: text,
+		Model: t.config.model,
+	}
+
+	requestBody, err := json.Marshal(data)
+	if err != nil {
+		log.Errorf("failed to marshal request data: %v", err)
+		return "", nil, nil, err
+	}
+
+	headers := [][2]string{
+		{"Authorization", fmt.Sprintf("Bearer %s", openAIConfig.apiKey)},
+		{"Content-Type", "application/json"},
+	}
+
+	return OPENAI_ENDPOINT, headers, requestBody, err
+}
+
+func (t *OpenAIProvider) parseTextEmbedding(responseBody []byte) (*OpenAIResponse, error) {
+	var resp OpenAIResponse
+	err := json.Unmarshal(responseBody, &resp)
+	if err != nil {
+		return nil, err
+	}
+	return &resp, nil
+}
+
+func (t *OpenAIProvider) GetEmbedding(
+	queryString string,
+	ctx wrapper.HttpContext,
+	log wrapper.Log,
+	callback func(emb []float64, err error)) error {
+	embUrl, embHeaders, embRequestBody, err := t.constructParameters(queryString, log)
+	if err != nil {
+		log.Errorf("failed to construct parameters: %v", err)
+		return err
+	}
+
+	var resp *OpenAIResponse
+	err = t.client.Post(embUrl, embHeaders, embRequestBody,
+		func(statusCode int, responseHeaders http.Header, responseBody []byte) {
+
+			if statusCode != http.StatusOK {
+				err = fmt.Errorf("failed to get embedding due to status code: %d, resp: %s", statusCode, responseBody)
+				callback(nil, err)
+				return
+			}
+
+			resp, err = t.parseTextEmbedding(responseBody)
+			if err != nil {
+				err = fmt.Errorf("failed to parse response: %v", err)
+				callback(nil, err)
+				return
+			}
+
+			log.Debugf("get embedding response: %d, %s", statusCode, responseBody)
+
+			if len(resp.Data) == 0 {
+				err = errors.New("no embedding found in response")
+				callback(nil, err)
+				return
+			}
+
+			callback(resp.Data[0].Embedding, nil)
+
+		}, t.config.timeout)
+	return err
+}
--- a/plugins/wasm-go/extensions/ai-cache/embedding/provider.go
+++ b/plugins/wasm-go/extensions/ai-cache/embedding/provider.go
@@ -10,10 +10,13 @@ import (
 const (
 	PROVIDER_TYPE_DASHSCOPE = "dashscope"
 	PROVIDER_TYPE_TEXTIN    = "textin"
+	PROVIDER_TYPE_COHERE    = "cohere"
+	PROVIDER_TYPE_OPENAI    = "openai"
 )

 type providerInitializer interface {
-	ValidateConfig(ProviderConfig) error
+	InitConfig(json gjson.Result)
+	ValidateConfig() error
 	CreateProvider(ProviderConfig) (Provider, error)
 }

@@ -21,6 +24,8 @@ var (
 	providerInitializers = map[string]providerInitializer{
 		PROVIDER_TYPE_DASHSCOPE: &dashScopeProviderInitializer{},
 		PROVIDER_TYPE_TEXTIN:    &textInProviderInitializer{},
+		PROVIDER_TYPE_COHERE:    &cohereProviderInitializer{},
+		PROVIDER_TYPE_OPENAI:    &openAIProviderInitializer{},
 	}
 )

@@ -37,35 +42,26 @@ type ProviderConfig struct {
 	// @Title zh-CN 文本特征提取服务端口
 	// @Description zh-CN 文本特征提取服务端口
 	servicePort int64
-	// @Title zh-CN 文本特征提取服务 API Key
-	// @Description zh-CN 文本特征提取服务 API Key
-	apiKey string
-	//@Title zh-CN TextIn x-ti-app-id
-	// @Description zh-CN 仅适用于 TextIn 服务。参考 https://www.textin.com/document/acge_text_embedding
-	textinAppId string
-	//@Title zh-CN TextIn x-ti-secret-code
-	// @Description zh-CN 仅适用于 TextIn 服务。参考 https://www.textin.com/document/acge_text_embedding
-	textinSecretCode string
-	//@Title zh-CN TextIn request matryoshka_dim
-	// @Description zh-CN 仅适用于 TextIn 服务, 指定返回的向量维度。参考 https://www.textin.com/document/acge_text_embedding
-	textinMatryoshkaDim int
 	// @Title zh-CN 文本特征提取服务超时时间
 	// @Description zh-CN 文本特征提取服务超时时间
 	timeout uint32
 	// @Title zh-CN 文本特征提取服务使用的模型
 	// @Description zh-CN 用于文本特征提取的模型名称, 在 DashScope 中默认为 "text-embedding-v1"
 	model string
+
+	initializer providerInitializer
 }

 func (c *ProviderConfig) FromJson(json gjson.Result) {
 	c.typ = json.Get("type").String()
+	i, has := providerInitializers[c.typ]
+	if has {
+		i.InitConfig(json)
+		c.initializer = i
+	}
 	c.serviceName = json.Get("serviceName").String()
 	c.serviceHost = json.Get("serviceHost").String()
 	c.servicePort = json.Get("servicePort").Int()
-	c.apiKey = json.Get("apiKey").String()
-	c.textinAppId = json.Get("textinAppId").String()
-	c.textinSecretCode = json.Get("textinSecretCode").String()
-	c.textinMatryoshkaDim = int(json.Get("textinMatryoshkaDim").Int())
 	c.timeout = uint32(json.Get("timeout").Int())
 	c.model = json.Get("model").String()
 	if c.timeout == 0 {
@@ -80,11 +76,10 @@ func (c *ProviderConfig) Validate() error {
 	if c.typ == "" {
 		return errors.New("embedding service type is required")
 	}
-	initializer, has := providerInitializers[c.typ]
-	if !has {
+	if c.initializer == nil {
 		return errors.New("unknown embedding service provider type: " + c.typ)
 	}
-	if err := initializer.ValidateConfig(*c); err != nil {
+	if err := c.initializer.ValidateConfig(); err != nil {
 		return err
 	}
 	return nil
--- a/plugins/wasm-go/extensions/ai-cache/embedding/textin.go
+++ b/plugins/wasm-go/extensions/ai-cache/embedding/textin.go
@@ -8,6 +8,7 @@ import (
 	"strconv"

 	"github.com/alibaba/higress/plugins/wasm-go/pkg/wrapper"
+	"github.com/tidwall/gjson"
 )

 const (
@@ -20,14 +21,34 @@ const (
 type textInProviderInitializer struct {
 }

-func (t *textInProviderInitializer) ValidateConfig(config ProviderConfig) error {
-	if config.textinAppId == "" {
-		return errors.New("embedding service TextIn App ID is required")
+var textInConfig textInProviderConfig
+
+type textInProviderConfig struct {
+	//@Title zh-CN TextIn x-ti-app-id
+	// @Description zh-CN 仅适用于 TextIn 服务。参考 https://www.textin.com/document/acge_text_embedding
+	textinAppId string
+	//@Title zh-CN TextIn x-ti-secret-code
+	// @Description zh-CN 仅适用于 TextIn 服务。参考 https://www.textin.com/document/acge_text_embedding
+	textinSecretCode string
+	//@Title zh-CN TextIn request matryoshka_dim
+	// @Description zh-CN 仅适用于 TextIn 服务, 指定返回的向量维度。参考 https://www.textin.com/document/acge_text_embedding
+	textinMatryoshkaDim int
+}
+
+func (c *textInProviderInitializer) InitConfig(json gjson.Result) {
+	textInConfig.textinAppId = json.Get("textinAppId").String()
+	textInConfig.textinSecretCode = json.Get("textinSecretCode").String()
+	textInConfig.textinMatryoshkaDim = int(json.Get("textinMatryoshkaDim").Int())
+}
+
+func (c *textInProviderInitializer) ValidateConfig() error {
+	if textInConfig.textinAppId == "" {
+		return errors.New("textinAppId is required")
 	}
-	if config.textinSecretCode == "" {
-		return errors.New("embedding service TextIn Secret Code is required")
+	if textInConfig.textinSecretCode == "" {
+		return errors.New("textinSecretCode is required")
 	}
-	if config.textinMatryoshkaDim == 0 {
+	if textInConfig.textinMatryoshkaDim == 0 {
 		return errors.New("embedding service TextIn Matryoshka Dim is required")
 	}
 	return nil
@@ -62,7 +83,7 @@ type TextInResponse struct {
 }

 type TextInResult struct {
-	Embeddings    [][]float64 `json:"embedding"` 
+	Embeddings    [][]float64 `json:"embedding"`
 	MatryoshkaDim int         `json:"matryoshka_dim"`
 }

@@ -80,7 +101,7 @@ func (t *TIProvider) constructParameters(texts []string, log wrapper.Log) (strin

 	data := TextInEmbeddingRequest{
 		Input:         texts,
-		MatryoshkaDim: t.config.textinMatryoshkaDim,
+		MatryoshkaDim: textInConfig.textinMatryoshkaDim,
 	}

 	requestBody, err := json.Marshal(data)
@@ -89,20 +110,20 @@ func (t *TIProvider) constructParameters(texts []string, log wrapper.Log) (strin
 		return "", nil, nil, err
 	}

-	if t.config.textinAppId == "" {
+	if textInConfig.textinAppId == "" {
 		err := errors.New("textinAppId is empty")
 		log.Errorf("failed to construct headers: %v", err)
 		return "", nil, nil, err
 	}
-	if t.config.textinSecretCode == "" {
+	if textInConfig.textinSecretCode == "" {
 		err := errors.New("textinSecretCode is empty")
 		log.Errorf("failed to construct headers: %v", err)
 		return "", nil, nil, err
 	}

 	headers := [][2]string{
-		{"x-ti-app-id", t.config.textinAppId},
-		{"x-ti-secret-code", t.config.textinSecretCode},
+		{"x-ti-app-id", textInConfig.textinAppId},
+		{"x-ti-secret-code", textInConfig.textinSecretCode},
 		{"Content-Type", "application/json"},
 	}

--- a/plugins/wasm-go/extensions/ai-cache/embedding/weaviate.go
+++ b/plugins/wasm-go/extensions/ai-cache/embedding/weaviate.go
@@ -1,27 +0,0 @@
-package embedding
-
-// import (
-// 	"github.com/alibaba/higress/plugins/wasm-go/pkg/wrapper"
-// )
-
-// const (
-// 	weaviateURL = "172.17.0.1:8081"
-// )
-
-// type weaviateProviderInitializer struct {
-// }
-
-// func (d *weaviateProviderInitializer) ValidateConfig(config ProviderConfig) error {
-// 	return nil
-// }
-
-// func (d *weaviateProviderInitializer) CreateProvider(config ProviderConfig) (Provider, error) {
-// 	return &DSProvider{
-// 		config: config,
-// 		client: wrapper.NewClusterClient(wrapper.DnsCluster{
-// 			ServiceName: config.ServiceName,
-// 			Port:        dashScopePort,
-// 			Domain:      dashScopeDomain,
-// 		}),
-// 	}, nil
-// }
--- a/plugins/wasm-go/extensions/ai-cache/go.mod
+++ b/plugins/wasm-go/extensions/ai-cache/go.mod
@@ -8,14 +8,14 @@ replace github.com/alibaba/higress/plugins/wasm-go => ../..

 require (
 	github.com/alibaba/higress/plugins/wasm-go v1.4.2
-	github.com/higress-group/proxy-wasm-go-sdk v0.0.0-20240711023527-ba358c48772f
+	github.com/google/uuid v1.6.0
+	github.com/higress-group/proxy-wasm-go-sdk v1.0.0
 	github.com/tidwall/gjson v1.17.3
 	github.com/tidwall/resp v0.1.1
 // github.com/weaviate/weaviate-go-client/v4 v4.15.1
 )

 require (
-	github.com/google/uuid v1.6.0 // indirect
 	github.com/higress-group/nottinygc v0.0.0-20231101025119-e93c4c2f8520 // indirect
 	github.com/magefile/mage v1.14.0 // indirect
 	github.com/stretchr/testify v1.9.0 // indirect
--- a/plugins/wasm-go/extensions/ai-cache/go.sum
+++ b/plugins/wasm-go/extensions/ai-cache/go.sum
@@ -3,8 +3,8 @@ github.com/google/uuid v1.6.0 h1:NIvaJDMOsjHA8n1jAhLSgzrAzy1Hgr+hNrb57e+94F0=
 github.com/google/uuid v1.6.0/go.mod h1:TIyPZe4MgqvfeYDBFedMoGGpEw/LqOeaOT+nhxU+yHo=
 github.com/higress-group/nottinygc v0.0.0-20231101025119-e93c4c2f8520 h1:IHDghbGQ2DTIXHBHxWfqCYQW1fKjyJ/I7W1pMyUDeEA=
 github.com/higress-group/nottinygc v0.0.0-20231101025119-e93c4c2f8520/go.mod h1:Nz8ORLaFiLWotg6GeKlJMhv8cci8mM43uEnLA5t8iew=
-github.com/higress-group/proxy-wasm-go-sdk v0.0.0-20240711023527-ba358c48772f h1:ZIiIBRvIw62gA5MJhuwp1+2wWbqL9IGElQ499rUsYYg=
-github.com/higress-group/proxy-wasm-go-sdk v0.0.0-20240711023527-ba358c48772f/go.mod h1:hNFjhrLUIq+kJ9bOcs8QtiplSQ61GZXtd2xHKx4BYRo=
+github.com/higress-group/proxy-wasm-go-sdk v1.0.0 h1:BZRNf4R7jr9hwRivg/E29nkVaKEak5MWjBDhWjuHijU=
+github.com/higress-group/proxy-wasm-go-sdk v1.0.0/go.mod h1:iiSyFbo+rAtbtGt/bsefv8GU57h9CCLYGJA74/tF5/0=
 github.com/magefile/mage v1.14.0 h1:6QDX3g6z1YvJ4olPhT1wksUcSa/V0a1B+pJb73fBjyo=
 github.com/magefile/mage v1.14.0/go.mod h1:z5UZb/iS3GoOSn0JgWuiw7dxlurVYTu+/jHXqQg881A=
 github.com/pmezard/go-difflib v1.0.0 h1:4DBwDE0NGyQoBHbLQYPwSUPoCMWR5BEzIk/f1lZbAQM=
--- a/plugins/wasm-go/extensions/ai-cache/main.go
+++ b/plugins/wasm-go/extensions/ai-cache/main.go
@@ -22,6 +22,8 @@ const (
 	STREAM_CONTEXT_KEY          = "stream"
 	SKIP_CACHE_HEADER           = "x-higress-skip-ai-cache"
 	ERROR_PARTIAL_MESSAGE_KEY   = "errorPartialMessage"
+
+	DEFAULT_MAX_BODY_BYTES uint32 = 10 * 1024 * 1024
 )

 func main() {
@@ -69,6 +71,7 @@ func onHttpRequestHeaders(ctx wrapper.HttpContext, c config.PluginConfig, log wr
 		ctx.DontReadRequestBody()
 		return types.ActionContinue
 	}
+	ctx.SetRequestBodyBufferLimit(DEFAULT_MAX_BODY_BYTES)
 	_ = proxywasm.RemoveHttpRequestHeader("Accept-Encoding")
 	// The request has a body and requires delaying the header transmission until a cache miss occurs,
 	// at which point the header should be sent.
@@ -128,12 +131,20 @@ func onHttpRequestBody(ctx wrapper.HttpContext, c config.PluginConfig, body []by
 func onHttpResponseHeaders(ctx wrapper.HttpContext, c config.PluginConfig, log wrapper.Log) types.Action {
 	skipCache := ctx.GetContext(SKIP_CACHE_HEADER)
 	if skipCache != nil {
+		ctx.SetUserAttribute("cache_status", "skip")
+		ctx.WriteUserAttributeToLogWithKey(wrapper.AILogKey)
 		ctx.DontReadResponseBody()
 		return types.ActionContinue
 	}
+	if ctx.GetContext(CACHE_KEY_CONTEXT_KEY) != nil {
+		ctx.SetUserAttribute("cache_status", "miss")
+		ctx.WriteUserAttributeToLogWithKey(wrapper.AILogKey)
+	}
 	contentType, _ := proxywasm.GetHttpResponseHeader("content-type")
 	if strings.Contains(contentType, "text/event-stream") {
 		ctx.SetContext(STREAM_CONTEXT_KEY, struct{}{})
+	} else {
+		ctx.SetResponseBodyBufferLimit(DEFAULT_MAX_BODY_BYTES)
 	}

 	if ctx.GetContext(ERROR_PARTIAL_MESSAGE_KEY) != nil {
@@ -158,22 +169,26 @@ func onHttpResponseBody(ctx wrapper.HttpContext, c config.PluginConfig, chunk []
 		return chunk
 	}

+	stream := ctx.GetContext(STREAM_CONTEXT_KEY)
+	var err error
 	if !isLastChunk {
-		if err := handleNonLastChunk(ctx, c, chunk, log); err != nil {
+		if stream == nil {
+			err = handleNonStreamChunk(ctx, c, chunk, log)
+		} else {
+			err = handleStreamChunk(ctx, c, unifySSEChunk(chunk), log)
+		}
+		if err != nil {
 			log.Errorf("[onHttpResponseBody] handle non last chunk failed, error: %v", err)
 			// Set an empty struct in the context to indicate an error in processing the partial message
 			ctx.SetContext(ERROR_PARTIAL_MESSAGE_KEY, struct{}{})
 		}
 		return chunk
 	}
-
-	stream := ctx.GetContext(STREAM_CONTEXT_KEY)
 	var value string
-	var err error
 	if stream == nil {
 		value, err = processNonStreamLastChunk(ctx, c, chunk, log)
 	} else {
-		value, err = processStreamLastChunk(ctx, c, chunk, log)
+		value, err = processStreamLastChunk(ctx, c, unifySSEChunk(chunk), log)
 	}

 	if err != nil {
--- a/plugins/wasm-go/extensions/ai-cache/util.go
+++ b/plugins/wasm-go/extensions/ai-cache/util.go
@@ -1,6 +1,7 @@
 package main

 import (
+	"bytes"
 	"fmt"
 	"strings"

@@ -9,17 +10,6 @@ import (
 	"github.com/tidwall/gjson"
 )

-func handleNonLastChunk(ctx wrapper.HttpContext, c config.PluginConfig, chunk []byte, log wrapper.Log) error {
-	stream := ctx.GetContext(STREAM_CONTEXT_KEY)
-	err := error(nil)
-	if stream == nil {
-		err = handleNonStreamChunk(ctx, c, chunk, log)
-	} else {
-		err = handleStreamChunk(ctx, c, chunk, log)
-	}
-	return err
-}
-
 func handleNonStreamChunk(ctx wrapper.HttpContext, c config.PluginConfig, chunk []byte, log wrapper.Log) error {
 	tempContentI := ctx.GetContext(CACHE_CONTENT_CONTEXT_KEY)
 	if tempContentI == nil {
@@ -32,6 +22,12 @@ func handleNonStreamChunk(ctx wrapper.HttpContext, c config.PluginConfig, chunk
 	return nil
 }

+func unifySSEChunk(data []byte) []byte {
+	data = bytes.ReplaceAll(data, []byte("\r\n"), []byte("\n"))
+	data = bytes.ReplaceAll(data, []byte("\r"), []byte("\n"))
+	return data
+}
+
 func handleStreamChunk(ctx wrapper.HttpContext, c config.PluginConfig, chunk []byte, log wrapper.Log) error {
 	var partialMessage []byte
 	partialMessageI := ctx.GetContext(PARTIAL_MESSAGE_CONTEXT_KEY)
@@ -101,55 +97,54 @@ func processStreamLastChunk(ctx wrapper.HttpContext, c config.PluginConfig, chun
 }

 func processSSEMessage(ctx wrapper.HttpContext, c config.PluginConfig, sseMessage string, log wrapper.Log) (string, error) {
-	subMessages := strings.Split(sseMessage, "\n")
-	var message string
-	for _, msg := range subMessages {
-		if strings.HasPrefix(msg, "data:") {
-			message = msg
-			break
+	content := ""
+	for _, chunk := range strings.Split(sseMessage, "\n\n") {
+		log.Debugf("single sse message: %s", chunk)
+		subMessages := strings.Split(chunk, "\n")
+		var message string
+		for _, msg := range subMessages {
+			if strings.HasPrefix(msg, "data:") {
+				message = msg
+				break
+			}
 		}
-	}
-	if len(message) < 6 {
-		return "", fmt.Errorf("[processSSEMessage] invalid message: %s", message)
-	}
-
-	// skip the prefix "data:"
-	bodyJson := message[5:]
-
-	if strings.TrimSpace(bodyJson) == "[DONE]" {
-		return "", nil
-	}
-
-	// Extract values from JSON fields
-	responseBody := gjson.Get(bodyJson, c.CacheStreamValueFrom)
-	toolCalls := gjson.Get(bodyJson, c.CacheToolCallsFrom)
-
-	if toolCalls.Exists() {
-		// TODO: Temporarily store the tool_calls value in the context for processing
-		ctx.SetContext(TOOL_CALLS_CONTEXT_KEY, toolCalls.String())
-	}
-
-	// Check if the ResponseBody field exists
-	if !responseBody.Exists() {
-		if ctx.GetContext(CACHE_CONTENT_CONTEXT_KEY) != nil {
-			log.Debugf("[processSSEMessage] unable to extract content from message; cache content is not nil: %s", message)
-			return "", nil
+		if len(message) < 6 {
+			return content, fmt.Errorf("[processSSEMessage] invalid message: %s", message)
 		}
-		return "", fmt.Errorf("[processSSEMessage] unable to extract content from message; cache content is nil: %s", message)
-	} else {
-		tempContentI := ctx.GetContext(CACHE_CONTENT_CONTEXT_KEY)

-		// If there is no content in the cache, initialize and set the content
-		if tempContentI == nil {
-			content := responseBody.String()
-			ctx.SetContext(CACHE_CONTENT_CONTEXT_KEY, content)
+		// skip the prefix "data:"
+		bodyJson := message[5:]
+
+		if strings.TrimSpace(bodyJson) == "[DONE]" {
 			return content, nil
 		}

-		// Update the content in the cache
-		appendMsg := responseBody.String()
-		content := tempContentI.(string) + appendMsg
-		ctx.SetContext(CACHE_CONTENT_CONTEXT_KEY, content)
-		return content, nil
+		// Extract values from JSON fields
+		responseBody := gjson.Get(bodyJson, c.CacheStreamValueFrom)
+		toolCalls := gjson.Get(bodyJson, c.CacheToolCallsFrom)
+
+		if toolCalls.Exists() {
+			// TODO: Temporarily store the tool_calls value in the context for processing
+			ctx.SetContext(TOOL_CALLS_CONTEXT_KEY, toolCalls.String())
+		}
+
+		// Check if the ResponseBody field exists
+		if !responseBody.Exists() {
+			if ctx.GetContext(CACHE_CONTENT_CONTEXT_KEY) != nil {
+				log.Debugf("[processSSEMessage] unable to extract content from message; cache content is not nil: %s", message)
+				return content, nil
+			}
+			return content, fmt.Errorf("[processSSEMessage] unable to extract content from message; cache content is nil: %s", message)
+		} else {
+			content += responseBody.String()
+		}
 	}
+	tempContentI := ctx.GetContext(CACHE_CONTENT_CONTEXT_KEY)
+	// If there is no content in the cache, initialize and set the content
+	if tempContentI == nil {
+		ctx.SetContext(CACHE_CONTENT_CONTEXT_KEY, content)
+	} else {
+		ctx.SetContext(CACHE_CONTENT_CONTEXT_KEY, tempContentI.(string)+content)
+	}
+	return content, nil
 }
--- a/plugins/wasm-go/extensions/ai-cache/vector/chroma.go
+++ b/plugins/wasm-go/extensions/ai-cache/vector/chroma.go
@@ -0,0 +1,201 @@
+package vector
+
+import (
+	"encoding/json"
+	"errors"
+	"fmt"
+	"net/http"
+
+	"github.com/alibaba/higress/plugins/wasm-go/pkg/wrapper"
+)
+
+type chromaProviderInitializer struct{}
+
+func (c *chromaProviderInitializer) ValidateConfig(config ProviderConfig) error {
+	if len(config.collectionID) == 0 {
+		return errors.New("[Chroma] collectionID is required")
+	}
+	if len(config.serviceName) == 0 {
+		return errors.New("[Chroma] serviceName is required")
+	}
+	return nil
+}
+
+func (c *chromaProviderInitializer) CreateProvider(config ProviderConfig) (Provider, error) {
+	return &ChromaProvider{
+		config: config,
+		client: wrapper.NewClusterClient(wrapper.FQDNCluster{
+			FQDN: config.serviceName,
+			Host: config.serviceHost,
+			Port: int64(config.servicePort),
+		}),
+	}, nil
+}
+
+type ChromaProvider struct {
+	config ProviderConfig
+	client wrapper.HttpClient
+}
+
+func (c *ChromaProvider) GetProviderType() string {
+	return PROVIDER_TYPE_CHROMA
+}
+
+func (d *ChromaProvider) QueryEmbedding(
+	emb []float64,
+	ctx wrapper.HttpContext,
+	log wrapper.Log,
+	callback func(results []QueryResult, ctx wrapper.HttpContext, log wrapper.Log, err error)) error {
+	// 最少需要填写的参数为 collection_id, embeddings 和 ids
+	// 下面是一个例子
+	// {
+	// 	"where": {}, // 用于 metadata 过滤，可选参数
+	// 	"where_document": {}, // 用于 document 过滤，可选参数
+	// 	"query_embeddings": [
+	// 	  [1.1, 2.3, 3.2]
+	// 	],
+	// 	"limit": 5,
+	// 	"include": [
+	// 	  "metadatas", // 可选
+	// 	  "documents", // 如果需要答案则需要
+	// 	  "distances"
+	// 	]
+	// }
+
+	requestBody, err := json.Marshal(chromaQueryRequest{
+		QueryEmbeddings: []chromaEmbedding{emb},
+		Limit:           d.config.topK,
+		Include:         []string{"distances", "documents"},
+	})
+
+	if err != nil {
+		log.Errorf("[Chroma] Failed to marshal query embedding request body: %v", err)
+		return err
+	}
+
+	return d.client.Post(
+		fmt.Sprintf("/api/v1/collections/%s/query", d.config.collectionID),
+		[][2]string{
+			{"Content-Type", "application/json"},
+		},
+		requestBody,
+		func(statusCode int, responseHeaders http.Header, responseBody []byte) {
+			log.Debugf("[Chroma] Query embedding response: %d, %s", statusCode, responseBody)
+			results, err := d.parseQueryResponse(responseBody, log)
+			if err != nil {
+				err = fmt.Errorf("[Chroma] Failed to parse query response: %v", err)
+			}
+			callback(results, ctx, log, err)
+		},
+		d.config.timeout,
+	)
+}
+
+func (d *ChromaProvider) UploadAnswerAndEmbedding(
+	queryString string,
+	queryEmb []float64,
+	queryAnswer string,
+	ctx wrapper.HttpContext,
+	log wrapper.Log,
+	callback func(ctx wrapper.HttpContext, log wrapper.Log, err error)) error {
+	// 最少需要填写的参数为 collection_id, embeddings 和 ids
+	// 下面是一个例子
+	// {
+	// 	"embeddings": [
+	// 		  [1.1, 2.3, 3.2]
+	// 	],
+	// 	"ids": [
+	// 	  "你吃了吗？"
+	// 	],
+	//  "documents": [
+	//    "我吃了。"
+	//  ]
+	// }
+	// 如果要添加 answer，则按照以下例子
+	// {
+	// 	"embeddings": [
+	// 	  [1.1, 2.3, 3.2]
+	// 	],
+	// 	"documents": [
+	// 	  "answer1"
+	// 	],
+	// 	"ids": [
+	// 	  "id1"
+	// 	]
+	// }
+	requestBody, err := json.Marshal(chromaInsertRequest{
+		Embeddings: []chromaEmbedding{queryEmb},
+		IDs:        []string{queryString}, // queryString 指的是用户查询的问题
+		Documents:  []string{queryAnswer}, // queryAnswer 指的是用户查询的问题的答案
+	})
+
+	if err != nil {
+		log.Errorf("[Chroma] Failed to marshal upload embedding request body: %v", err)
+		return err
+	}
+
+	err = d.client.Post(
+		fmt.Sprintf("/api/v1/collections/%s/add", d.config.collectionID),
+		[][2]string{
+			{"Content-Type", "application/json"},
+		},
+		requestBody,
+		func(statusCode int, responseHeaders http.Header, responseBody []byte) {
+			log.Debugf("[Chroma] statusCode:%d, responseBody:%s", statusCode, string(responseBody))
+			callback(ctx, log, err)
+		},
+		d.config.timeout,
+	)
+	return err
+}
+
+type chromaEmbedding []float64
+type chromaMetadataMap map[string]string
+type chromaInsertRequest struct {
+	Embeddings []chromaEmbedding   `json:"embeddings"`
+	Metadatas  []chromaMetadataMap `json:"metadatas,omitempty"` // 可选参数
+	Documents  []string            `json:"documents,omitempty"` // 可选参数
+	IDs        []string            `json:"ids"`
+}
+
+type chromaQueryRequest struct {
+	Where           map[string]string `json:"where,omitempty"`          // 可选参数
+	WhereDocument   map[string]string `json:"where_document,omitempty"` // 可选参数
+	QueryEmbeddings []chromaEmbedding `json:"query_embeddings"`
+	Limit           int               `json:"limit"`
+	Include         []string          `json:"include"`
+}
+
+type chromaQueryResponse struct {
+	Ids        [][]string          `json:"ids"`                  // 第一维是 batch query，第二维是查询到的多个 ids
+	Distances  [][]float64         `json:"distances,omitempty"`  // 与 Ids 一一对应
+	Metadatas  []chromaMetadataMap `json:"metadatas,omitempty"`  // 可选参数
+	Embeddings []chromaEmbedding   `json:"embeddings,omitempty"` // 可选参数
+	Documents  [][]string          `json:"documents,omitempty"`  // 与 Ids 一一对应
+	Uris       []string            `json:"uris,omitempty"`       // 可选参数
+	Data       []interface{}       `json:"data,omitempty"`       // 可选参数
+	Included   []string            `json:"included"`
+}
+
+func (d *ChromaProvider) parseQueryResponse(responseBody []byte, log wrapper.Log) ([]QueryResult, error) {
+	var queryResp chromaQueryResponse
+	err := json.Unmarshal(responseBody, &queryResp)
+	if err != nil {
+		return nil, err
+	}
+
+	log.Debugf("[Chroma] queryResp Ids len: %d", len(queryResp.Ids))
+	if len(queryResp.Ids) == 1 && len(queryResp.Ids[0]) == 0 {
+		return nil, errors.New("no query results found in response")
+	}
+	results := make([]QueryResult, 0, len(queryResp.Ids[0]))
+	for i := range queryResp.Ids[0] {
+		result := QueryResult{
+			Text:   queryResp.Ids[0][i],
+			Score:  queryResp.Distances[0][i],
+			Answer: queryResp.Documents[0][i],
+		}
+		results = append(results, result)
+	}
+	return results, nil
+}
--- a/plugins/wasm-go/extensions/ai-cache/vector/elasticsearch.go
+++ b/plugins/wasm-go/extensions/ai-cache/vector/elasticsearch.go
@@ -0,0 +1,200 @@
+package vector
+
+import (
+	"encoding/base64"
+	"encoding/json"
+	"errors"
+	"fmt"
+	"net/http"
+
+	"github.com/alibaba/higress/plugins/wasm-go/pkg/wrapper"
+)
+
+type esProviderInitializer struct{}
+
+func (c *esProviderInitializer) ValidateConfig(config ProviderConfig) error {
+	if len(config.collectionID) == 0 {
+		return errors.New("[ES] collectionID is required")
+	}
+	if len(config.serviceName) == 0 {
+		return errors.New("[ES] serviceName is required")
+	}
+	return nil
+}
+
+func (c *esProviderInitializer) CreateProvider(config ProviderConfig) (Provider, error) {
+	return &ESProvider{
+		config: config,
+		client: wrapper.NewClusterClient(wrapper.FQDNCluster{
+			FQDN: config.serviceName,
+			Host: config.serviceHost,
+			Port: int64(config.servicePort),
+		}),
+	}, nil
+}
+
+type ESProvider struct {
+	config ProviderConfig
+	client wrapper.HttpClient
+}
+
+func (c *ESProvider) GetProviderType() string {
+	return PROVIDER_TYPE_ES
+}
+
+func (d *ESProvider) QueryEmbedding(
+	emb []float64,
+	ctx wrapper.HttpContext,
+	log wrapper.Log,
+	callback func(results []QueryResult, ctx wrapper.HttpContext, log wrapper.Log, err error)) error {
+
+	requestBody, err := json.Marshal(esQueryRequest{
+		Source: Source{Excludes: []string{"embedding"}},
+		Knn: knn{
+			Field:       "embedding",
+			QueryVector: emb,
+			K:           d.config.topK,
+		},
+		Size: d.config.topK,
+	})
+
+	if err != nil {
+		log.Errorf("[ES] Failed to marshal query embedding request body: %v", err)
+		return err
+	}
+
+	return d.client.Post(
+		fmt.Sprintf("/%s/_search", d.config.collectionID),
+		[][2]string{
+			{"Content-Type", "application/json"},
+			{"Authorization", d.getCredentials()},
+		},
+		requestBody,
+		func(statusCode int, responseHeaders http.Header, responseBody []byte) {
+			log.Debugf("[ES] Query embedding response: %d, %s", statusCode, responseBody)
+			results, err := d.parseQueryResponse(responseBody, log)
+			if err != nil {
+				err = fmt.Errorf("[ES] Failed to parse query response: %v", err)
+			}
+			callback(results, ctx, log, err)
+		},
+		d.config.timeout,
+	)
+}
+
+// base64 编码 ES 身份认证字符串或使用 Apikey
+func (d *ESProvider) getCredentials() string {
+	if len(d.config.apiKey) != 0 {
+		return fmt.Sprintf("ApiKey %s", d.config.apiKey)
+	} else {
+		credentials := fmt.Sprintf("%s:%s", d.config.esUsername, d.config.esPassword)
+		encodedCredentials := base64.StdEncoding.EncodeToString([]byte(credentials))
+		return fmt.Sprintf("Basic %s", encodedCredentials)
+	}
+
+}
+
+func (d *ESProvider) UploadAnswerAndEmbedding(
+	queryString string,
+	queryEmb []float64,
+	queryAnswer string,
+	ctx wrapper.HttpContext,
+	log wrapper.Log,
+	callback func(ctx wrapper.HttpContext, log wrapper.Log, err error)) error {
+	// 最少需要填写的参数为 index, embeddings 和 question
+	// 下面是一个例子
+	// POST /<index>/_doc
+	// {
+	// 	"embedding": [
+	// 		  [1.1, 2.3, 3.2]
+	// 	],
+	// 	"question": [
+	// 	  "你吃了吗？"
+	// 	]
+	// }
+	requestBody, err := json.Marshal(esInsertRequest{
+		Embedding: queryEmb,
+		Question:  queryString,
+		Answer:    queryAnswer,
+	})
+	if err != nil {
+		log.Errorf("[ES] Failed to marshal upload embedding request body: %v", err)
+		return err
+	}
+
+	return d.client.Post(
+		fmt.Sprintf("/%s/_doc", d.config.collectionID),
+		[][2]string{
+			{"Content-Type", "application/json"},
+			{"Authorization", d.getCredentials()},
+		},
+		requestBody,
+		func(statusCode int, responseHeaders http.Header, responseBody []byte) {
+			log.Debugf("[ES] statusCode:%d, responseBody:%s", statusCode, string(responseBody))
+			callback(ctx, log, err)
+		},
+		d.config.timeout,
+	)
+}
+
+type esInsertRequest struct {
+	Embedding []float64 `json:"embedding"`
+	Question  string    `json:"question"`
+	Answer    string    `json:"answer"`
+}
+
+type knn struct {
+	Field       string    `json:"field"`
+	QueryVector []float64 `json:"query_vector"`
+	K           int       `json:"k"`
+}
+
+type Source struct {
+	Excludes []string `json:"excludes"`
+}
+
+type esQueryRequest struct {
+	Source Source `json:"_source"`
+	Knn    knn    `json:"knn"`
+	Size   int    `json:"size"`
+}
+
+type esQueryResponse struct {
+	Took     int  `json:"took"`
+	TimedOut bool `json:"timed_out"`
+	Hits     struct {
+		Total struct {
+			Value    int    `json:"value"`
+			Relation string `json:"relation"`
+		} `json:"total"`
+		Hits []struct {
+			Index  string                 `json:"_index"`
+			ID     string                 `json:"_id"`
+			Score  float64                `json:"_score"`
+			Source map[string]interface{} `json:"_source"`
+		} `json:"hits"`
+	} `json:"hits"`
+}
+
+func (d *ESProvider) parseQueryResponse(responseBody []byte, log wrapper.Log) ([]QueryResult, error) {
+	log.Infof("[ES] responseBody: %s", string(responseBody))
+	var queryResp esQueryResponse
+	err := json.Unmarshal(responseBody, &queryResp)
+	if err != nil {
+		return []QueryResult{}, err
+	}
+	log.Debugf("[ES] queryResp Hits len: %d", len(queryResp.Hits.Hits))
+	if len(queryResp.Hits.Hits) == 0 {
+		return nil, errors.New("no query results found in response")
+	}
+	results := make([]QueryResult, 0, queryResp.Hits.Total.Value)
+	for _, hit := range queryResp.Hits.Hits {
+		result := QueryResult{
+			Text:   hit.Source["question"].(string),
+			Score:  hit.Score,
+			Answer: hit.Source["answer"].(string),
+		}
+		results = append(results, result)
+	}
+	return results, nil
+}
--- a/plugins/wasm-go/extensions/ai-cache/vector/milvus.go
+++ b/plugins/wasm-go/extensions/ai-cache/vector/milvus.go
@@ -0,0 +1,206 @@
+package vector
+
+import (
+	"encoding/json"
+	"errors"
+	"fmt"
+	"net/http"
+
+	"github.com/alibaba/higress/plugins/wasm-go/pkg/wrapper"
+	"github.com/tidwall/gjson"
+)
+
+type milvusProviderInitializer struct{}
+
+func (c *milvusProviderInitializer) ValidateConfig(config ProviderConfig) error {
+	if len(config.serviceName) == 0 {
+		return errors.New("[Milvus] serviceName is required")
+	}
+	if len(config.collectionID) == 0 {
+		return errors.New("[Milvus] collectionID is required")
+	}
+	return nil
+}
+
+func (c *milvusProviderInitializer) CreateProvider(config ProviderConfig) (Provider, error) {
+	return &milvusProvider{
+		config: config,
+		client: wrapper.NewClusterClient(wrapper.FQDNCluster{
+			FQDN: config.serviceName,
+			Host: config.serviceHost,
+			Port: int64(config.servicePort),
+		}),
+	}, nil
+}
+
+type milvusProvider struct {
+	config ProviderConfig
+	client wrapper.HttpClient
+}
+
+func (c *milvusProvider) GetProviderType() string {
+	return PROVIDER_TYPE_MILVUS
+}
+
+type milvusData struct {
+	Vector   []float64 `json:"vector"`
+	Question string    `json:"question,omitempty"`
+	Answer   string    `json:"answer,omitempty"`
+}
+
+type milvusInsertRequest struct {
+	CollectionName string       `json:"collectionName"`
+	Data           []milvusData `json:"data"`
+}
+
+func (d *milvusProvider) UploadAnswerAndEmbedding(
+	queryString string,
+	queryEmb []float64,
+	queryAnswer string,
+	ctx wrapper.HttpContext,
+	log wrapper.Log,
+	callback func(ctx wrapper.HttpContext, log wrapper.Log, err error)) error {
+	// 最少需要填写的参数为 collectionName, data 和 Authorization. question, answer 可选
+	// 需要填写 id，否则 v2.4.13-hotfix 提示 invalid syntax: invalid parameter[expected=Int64][actual=]
+	// 如果不填写 id，要在创建 collection 的时候设置 autoId 为 true
+	// 下面是一个例子
+	// {
+	// 	"collectionName": "higress",
+	// 	"data": [
+	// 	  {
+	// 	    "question": "这里是问题",
+	// 	  	"answer": "这里是答案"
+	// 	    "vector": [
+	// 	      0.9,
+	// 	      0.1,
+	// 	      0.1
+	// 	    ]
+	// 	  }
+	//   ]
+	// }
+	requestBody, err := json.Marshal(milvusInsertRequest{
+		CollectionName: d.config.collectionID,
+		Data: []milvusData{
+			{
+				Question: queryString,
+				Answer:   queryAnswer,
+				Vector:   queryEmb,
+			},
+		},
+	})
+
+	if err != nil {
+		log.Errorf("[Milvus] Failed to marshal upload embedding request body: %v", err)
+		return err
+	}
+
+	return d.client.Post(
+		"/v2/vectordb/entities/insert",
+		[][2]string{
+			{"Content-Type", "application/json"},
+			{"Authorization", fmt.Sprintf("Bearer %s", d.config.apiKey)},
+		},
+		requestBody,
+		func(statusCode int, responseHeaders http.Header, responseBody []byte) {
+			log.Debugf("[Milvus] statusCode:%d, responseBody:%s", statusCode, string(responseBody))
+			callback(ctx, log, err)
+		},
+		d.config.timeout,
+	)
+}
+
+type milvusQueryRequest struct {
+	CollectionName string      `json:"collectionName"`
+	Data           [][]float64 `json:"data"`
+	AnnsField      string      `json:"annsField"`
+	Limit          int         `json:"limit"`
+	OutputFields   []string    `json:"outputFields"`
+}
+
+func (d *milvusProvider) QueryEmbedding(
+	emb []float64,
+	ctx wrapper.HttpContext,
+	log wrapper.Log,
+	callback func(results []QueryResult, ctx wrapper.HttpContext, log wrapper.Log, err error)) error {
+	// 最少需要填写的参数为 collectionName, data, annsField. outputFields 为可选参数
+	// 下面是一个例子
+	// {
+	// 	"collectionName": "quick_setup",
+	// 	"data": [
+	// 		[
+	// 			0.3580376395471989,
+	// 			"Unknown type",
+	// 			0.18414012509913835,
+	// 			"Unknown type",
+	// 			0.9029438446296592
+	// 		]
+	// 	],
+	// 	"annsField": "vector",
+	// 	"limit": 3,
+	// 	"outputFields": [
+	// 		"color"
+	// 	]
+	// }
+	requestBody, err := json.Marshal(milvusQueryRequest{
+		CollectionName: d.config.collectionID,
+		Data:           [][]float64{emb},
+		AnnsField:      "vector",
+		Limit:          d.config.topK,
+		OutputFields: []string{
+			"question",
+			"answer",
+		},
+	})
+	if err != nil {
+		log.Errorf("[Milvus] Failed to marshal query embedding: %v", err)
+		return err
+	}
+
+	return d.client.Post(
+		"/v2/vectordb/entities/search",
+		[][2]string{
+			{"Content-Type", "application/json"},
+			{"Authorization", fmt.Sprintf("Bearer %s", d.config.apiKey)},
+		},
+		requestBody,
+		func(statusCode int, responseHeaders http.Header, responseBody []byte) {
+			log.Debugf("[Milvus] Query embedding response: %d, %s", statusCode, responseBody)
+			results, err := d.parseQueryResponse(responseBody, log)
+			if err != nil {
+				err = fmt.Errorf("[Milvus] Failed to parse query response: %v", err)
+			}
+			callback(results, ctx, log, err)
+		},
+		d.config.timeout,
+	)
+}
+
+func (d *milvusProvider) parseQueryResponse(responseBody []byte, log wrapper.Log) ([]QueryResult, error) {
+	if !gjson.GetBytes(responseBody, "data.0.distance").Exists() {
+		log.Errorf("[Milvus] No distance found in response body: %s", responseBody)
+		return nil, errors.New("[Milvus] No distance found in response body")
+	}
+
+	if !gjson.GetBytes(responseBody, "data.0.question").Exists() {
+		log.Errorf("[Milvus] No question found in response body: %s", responseBody)
+		return nil, errors.New("[Milvus] No question found in response body")
+	}
+
+	if !gjson.GetBytes(responseBody, "data.0.answer").Exists() {
+		log.Errorf("[Milvus] No answer found in response body: %s", responseBody)
+		return nil, errors.New("[Milvus] No answer found in response body")
+	}
+
+	resultNum := gjson.GetBytes(responseBody, "data.#").Int()
+	results := make([]QueryResult, 0, resultNum)
+	for i := 0; i < int(resultNum); i++ {
+		result := QueryResult{
+			Text:   gjson.GetBytes(responseBody, fmt.Sprintf("data.%d.question", i)).String(),
+			Score:  gjson.GetBytes(responseBody, fmt.Sprintf("data.%d.distance", i)).Float(),
+			Answer: gjson.GetBytes(responseBody, fmt.Sprintf("data.%d.answer", i)).String(),
+		}
+		results = append(results, result)
+	}
+
+	return results, nil
+}
--- a/plugins/wasm-go/extensions/ai-cache/vector/pinecone.go
+++ b/plugins/wasm-go/extensions/ai-cache/vector/pinecone.go
@@ -0,0 +1,194 @@
+package vector
+
+import (
+	"encoding/json"
+	"errors"
+	"fmt"
+	"net/http"
+
+	"github.com/alibaba/higress/plugins/wasm-go/pkg/wrapper"
+	"github.com/google/uuid"
+	"github.com/tidwall/gjson"
+)
+
+type pineconeProviderInitializer struct{}
+
+func (c *pineconeProviderInitializer) ValidateConfig(config ProviderConfig) error {
+	if len(config.serviceHost) == 0 {
+		return errors.New("[Pinecone] serviceHost is required")
+	}
+	if len(config.serviceName) == 0 {
+		return errors.New("[Pinecone] serviceName is required")
+	}
+	if len(config.apiKey) == 0 {
+		return errors.New("[Pinecone] apiKey is required")
+	}
+	return nil
+}
+
+func (c *pineconeProviderInitializer) CreateProvider(config ProviderConfig) (Provider, error) {
+	return &pineconeProvider{
+		config: config,
+		client: wrapper.NewClusterClient(wrapper.FQDNCluster{
+			FQDN: config.serviceName,
+			Host: config.serviceHost,
+			Port: int64(config.servicePort),
+		}),
+	}, nil
+}
+
+type pineconeProvider struct {
+	config ProviderConfig
+	client wrapper.HttpClient
+}
+
+func (c *pineconeProvider) GetProviderType() string {
+	return PROVIDER_TYPE_PINECONE
+}
+
+type pineconeMetadata struct {
+	Question string `json:"question"`
+	Answer   string `json:"answer"`
+}
+
+type pineconeVector struct {
+	ID         string           `json:"id"`
+	Values     []float64        `json:"values"`
+	Properties pineconeMetadata `json:"metadata"`
+}
+
+type pineconeInsertRequest struct {
+	Vectors   []pineconeVector `json:"vectors"`
+	Namespace string           `json:"namespace"`
+}
+
+func (d *pineconeProvider) UploadAnswerAndEmbedding(
+	queryString string,
+	queryEmb []float64,
+	queryAnswer string,
+	ctx wrapper.HttpContext,
+	log wrapper.Log,
+	callback func(ctx wrapper.HttpContext, log wrapper.Log, err error)) error {
+	// 最少需要填写的参数为 vector 和 question
+	// 下面是一个例子
+	// {
+	// 	"vectors": [
+	// 	  {
+	// 		"id": "A",
+	// 		"values": [0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1],
+	// 		"metadata": {"question": "你好", "answer": "你也好"}
+	// 	  }
+	// 	]
+	// }
+	requestBody, err := json.Marshal(pineconeInsertRequest{
+		Vectors: []pineconeVector{
+			{
+				ID:         uuid.New().String(),
+				Values:     queryEmb,
+				Properties: pineconeMetadata{Question: queryString, Answer: queryAnswer},
+			},
+		},
+		Namespace: d.config.collectionID,
+	})
+
+	if err != nil {
+		log.Errorf("[Pinecone] Failed to marshal upload embedding request body: %v", err)
+		return err
+	}
+
+	return d.client.Post(
+		"/vectors/upsert",
+		[][2]string{
+			{"Content-Type", "application/json"},
+			{"Api-Key", d.config.apiKey},
+		},
+		requestBody,
+		func(statusCode int, responseHeaders http.Header, responseBody []byte) {
+			log.Debugf("[Pinecone] statusCode:%d, responseBody:%s", statusCode, string(responseBody))
+			callback(ctx, log, err)
+		},
+		d.config.timeout,
+	)
+}
+
+type pineconeQueryRequest struct {
+	Namespace       string    `json:"namespace"`
+	Vector          []float64 `json:"vector"`
+	TopK            int       `json:"topK"`
+	IncludeMetadata bool      `json:"includeMetadata"`
+	IncludeValues   bool      `json:"includeValues"`
+}
+
+func (d *pineconeProvider) QueryEmbedding(
+	emb []float64,
+	ctx wrapper.HttpContext,
+	log wrapper.Log,
+	callback func(results []QueryResult, ctx wrapper.HttpContext, log wrapper.Log, err error)) error {
+	// 最少需要填写的参数为 vector
+	// 下面是一个例子
+	// {
+	// 	"namespace": "higress",
+	// 	"vector": [0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1],
+	// 	"topK": 1,
+	// 	"includeMetadata": false
+	// }
+	requestBody, err := json.Marshal(pineconeQueryRequest{
+		Namespace:       d.config.collectionID,
+		Vector:          emb,
+		TopK:            d.config.topK,
+		IncludeMetadata: true,
+		IncludeValues:   false,
+	})
+	if err != nil {
+		log.Errorf("[Pinecone] Failed to marshal query embedding: %v", err)
+		return err
+	}
+
+	return d.client.Post(
+		"/query",
+		[][2]string{
+			{"Content-Type", "application/json"},
+			{"Api-Key", d.config.apiKey},
+		},
+		requestBody,
+		func(statusCode int, responseHeaders http.Header, responseBody []byte) {
+			log.Debugf("[Pinecone] Query embedding response: %d, %s", statusCode, responseBody)
+			results, err := d.parseQueryResponse(responseBody, log)
+			if err != nil {
+				err = fmt.Errorf("[Pinecone] Failed to parse query response: %v", err)
+			}
+			callback(results, ctx, log, err)
+		},
+		d.config.timeout,
+	)
+}
+
+func (d *pineconeProvider) parseQueryResponse(responseBody []byte, log wrapper.Log) ([]QueryResult, error) {
+	if !gjson.GetBytes(responseBody, "matches.0.score").Exists() {
+		log.Errorf("[Pinecone] No distance found in response body: %s", responseBody)
+		return nil, errors.New("[Pinecone] No distance found in response body")
+	}
+
+	if !gjson.GetBytes(responseBody, "matches.0.metadata.question").Exists() {
+		log.Errorf("[Pinecone] No question found in response body: %s", responseBody)
+		return nil, errors.New("[Pinecone] No question found in response body")
+	}
+
+	if !gjson.GetBytes(responseBody, "matches.0.metadata.answer").Exists() {
+		log.Errorf("[Pinecone] No answer found in response body: %s", responseBody)
+		return nil, errors.New("[Pinecone] No answer found in response body")
+	}
+
+	resultNum := gjson.GetBytes(responseBody, "matches.#").Int()
+	results := make([]QueryResult, 0, resultNum)
+	for i := 0; i < int(resultNum); i++ {
+		result := QueryResult{
+			Text:   gjson.GetBytes(responseBody, fmt.Sprintf("matches.%d.metadata.question", i)).String(),
+			Score:  gjson.GetBytes(responseBody, fmt.Sprintf("matches.%d.score", i)).Float(),
+			Answer: gjson.GetBytes(responseBody, fmt.Sprintf("matches.%d.metadata.answer", i)).String(),
+		}
+		results = append(results, result)
+	}
+
+	return results, nil
+}
--- a/plugins/wasm-go/extensions/ai-cache/vector/provider.go
+++ b/plugins/wasm-go/extensions/ai-cache/vector/provider.go
@@ -10,6 +10,11 @@ import (
 const (
 	PROVIDER_TYPE_DASH_VECTOR = "dashvector"
 	PROVIDER_TYPE_CHROMA      = "chroma"
+	PROVIDER_TYPE_ES          = "elasticsearch"
+	PROVIDER_TYPE_WEAVIATE    = "weaviate"
+	PROVIDER_TYPE_PINECONE    = "pinecone"
+	PROVIDER_TYPE_QDRANT      = "qdrant"
+	PROVIDER_TYPE_MILVUS      = "milvus"
 )

 type providerInitializer interface {
@@ -20,7 +25,12 @@ type providerInitializer interface {
 var (
 	providerInitializers = map[string]providerInitializer{
 		PROVIDER_TYPE_DASH_VECTOR: &dashVectorProviderInitializer{},
-		// PROVIDER_TYPE_CHROMA:     &chromaProviderInitializer{},
+		PROVIDER_TYPE_CHROMA:      &chromaProviderInitializer{},
+		PROVIDER_TYPE_ES:          &esProviderInitializer{},
+		PROVIDER_TYPE_WEAVIATE:    &weaviateProviderInitializer{},
+		PROVIDER_TYPE_PINECONE:    &pineconeProviderInitializer{},
+		PROVIDER_TYPE_QDRANT:      &qdrantProviderInitializer{},
+		PROVIDER_TYPE_MILVUS:      &milvusProviderInitializer{},
 	}
 )

@@ -71,10 +81,6 @@ type StringQuerier interface {
 		callback func(results []QueryResult, ctx wrapper.HttpContext, log wrapper.Log, err error)) error
 }

-type SimilarityThresholdProvider interface {
-	GetSimilarityThreshold() float64
-}
-
 type ProviderConfig struct {
 	// @Title zh-CN 向量存储服务提供者类型
 	// @Description zh-CN 向量存储服务提供者类型，例如 dashvector、chroma
@@ -97,8 +103,8 @@ type ProviderConfig struct {
 	// @Title zh-CN 请求超时
 	// @Description zh-CN 请求向量存储服务的超时时间，单位为毫秒。默认值是10000，即10秒
 	timeout uint32
-	// @Title zh-CN DashVector 向量存储服务 Collection ID
-	// @Description zh-CN DashVector 向量存储服务 Collection ID
+	// @Title zh-CN 向量存储服务 Collection ID
+	// @Description zh-CN 向量存储服务的 Collection ID
 	collectionID string
 	// @Title zh-CN 相似度度量阈值
 	// @Description zh-CN 默认相似度度量阈值，默认为 1000。
@@ -109,6 +115,14 @@ type ProviderConfig struct {
 	// 所以需要允许自定义比较方式，对于 Cosine 和 DotProduct 选择 gt，对于 Euclidean 则选择 lt。
 	// 默认为 lt，所有条件包括 lt (less than，小于)、lte (less than or equal to，小等于)、gt (greater than，大于)、gte (greater than or equal to，大等于)
 	ThresholdRelation string
+
+	// ES 配置
+	// @Title zh-CN ES 用户名
+	// @Description zh-CN ES 用户名
+	esUsername string
+	// @Title zh-CN ES 密码
+	// @Description zh-CN ES 密码
+	esPassword string
 }

 func (c *ProviderConfig) GetProviderType() string {
@@ -117,7 +131,6 @@ func (c *ProviderConfig) GetProviderType() string {

 func (c *ProviderConfig) FromJson(json gjson.Result) {
 	c.typ = json.Get("type").String()
-	// DashVector
 	c.serviceName = json.Get("serviceName").String()
 	c.serviceHost = json.Get("serviceHost").String()
 	c.servicePort = int64(json.Get("servicePort").Int())
@@ -142,6 +155,10 @@ func (c *ProviderConfig) FromJson(json gjson.Result) {
 	if c.ThresholdRelation == "" {
 		c.ThresholdRelation = "lt"
 	}
+
+	// ES
+	c.esUsername = json.Get("esUsername").String()
+	c.esPassword = json.Get("esPassword").String()
 }

 func (c *ProviderConfig) Validate() error {
@@ -152,6 +169,9 @@ func (c *ProviderConfig) Validate() error {
 	if !has {
 		return errors.New("unknown vector database service provider type: " + c.typ)
 	}
+	if !isRelationValid(c.ThresholdRelation) {
+		return errors.New("invalid thresholdRelation: " + c.ThresholdRelation)
+	}
 	if err := initializer.ValidateConfig(*c); err != nil {
 		return err
 	}
@@ -165,3 +185,12 @@ func CreateProvider(pc ProviderConfig) (Provider, error) {
 	}
 	return initializer.CreateProvider(pc)
 }
+
+func isRelationValid(relation string) bool {
+	for _, r := range []string{"lt", "lte", "gt", "gte"} {
+		if r == relation {
+			return true
+		}
+	}
+	return false
+}
--- a/plugins/wasm-go/extensions/ai-cache/vector/qdrant.go
+++ b/plugins/wasm-go/extensions/ai-cache/vector/qdrant.go
@@ -0,0 +1,208 @@
+package vector
+
+import (
+	"encoding/json"
+	"errors"
+	"fmt"
+	"net/http"
+
+	"github.com/alibaba/higress/plugins/wasm-go/pkg/wrapper"
+	"github.com/google/uuid"
+	"github.com/tidwall/gjson"
+)
+
+type qdrantProviderInitializer struct{}
+
+func (c *qdrantProviderInitializer) ValidateConfig(config ProviderConfig) error {
+	if len(config.serviceName) == 0 {
+		return errors.New("[Qdrant] serviceName is required")
+	}
+	if len(config.collectionID) == 0 {
+		return errors.New("[Qdrant] collectionID is required")
+	}
+	return nil
+}
+
+func (c *qdrantProviderInitializer) CreateProvider(config ProviderConfig) (Provider, error) {
+	return &qdrantProvider{
+		config: config,
+		client: wrapper.NewClusterClient(wrapper.FQDNCluster{
+			FQDN: config.serviceName,
+			Host: config.serviceHost,
+			Port: int64(config.servicePort),
+		}),
+	}, nil
+}
+
+type qdrantProvider struct {
+	config ProviderConfig
+	client wrapper.HttpClient
+}
+
+func (c *qdrantProvider) GetProviderType() string {
+	return PROVIDER_TYPE_QDRANT
+}
+
+type qdrantPayload struct {
+	Question string `json:"question"`
+	Answer   string `json:"answer"`
+}
+
+type qdrantPoint struct {
+	ID      string        `json:"id"`
+	Vector  []float64     `json:"vector"`
+	Payload qdrantPayload `json:"payload"`
+}
+
+type qdrantInsertRequest struct {
+	Points []qdrantPoint `json:"points"`
+}
+
+func (d *qdrantProvider) UploadAnswerAndEmbedding(
+	queryString string,
+	queryEmb []float64,
+	queryAnswer string,
+	ctx wrapper.HttpContext,
+	log wrapper.Log,
+	callback func(ctx wrapper.HttpContext, log wrapper.Log, err error)) error {
+	// 最少需要填写的参数为 id 和 vector. payload 可选
+	// 下面是一个例子
+	// {
+	// 	"points": [
+	// 	  {
+	// 	    "id": "76874cce-1fb9-4e16-9b0b-f085ac06ed6f",
+	// 	    "payload": {
+	// 	      "question": "这里是问题",
+	// 	  	  "answer": "这里是答案"
+	// 	    },
+	// 	    "vector": [
+	// 	      0.9,
+	// 	      0.1,
+	// 	      0.1
+	// 	    ]
+	// 	  }
+	//   ]
+	// }
+	requestBody, err := json.Marshal(qdrantInsertRequest{
+		Points: []qdrantPoint{
+			{
+				ID:      uuid.New().String(),
+				Vector:  queryEmb,
+				Payload: qdrantPayload{Question: queryString, Answer: queryAnswer},
+			},
+		},
+	})
+
+	if err != nil {
+		log.Errorf("[Qdrant] Failed to marshal upload embedding request body: %v", err)
+		return err
+	}
+
+	return d.client.Put(
+		fmt.Sprintf("/collections/%s/points", d.config.collectionID),
+		[][2]string{
+			{"Content-Type", "application/json"},
+			{"api-key", d.config.apiKey},
+		},
+		requestBody,
+		func(statusCode int, responseHeaders http.Header, responseBody []byte) {
+			log.Debugf("[Qdrant] statusCode:%d, responseBody:%s", statusCode, string(responseBody))
+			callback(ctx, log, err)
+		},
+		d.config.timeout,
+	)
+}
+
+type qdrantQueryRequest struct {
+	Vector      []float64 `json:"vector"`
+	Limit       int       `json:"limit"`
+	WithPayload bool      `json:"with_payload"`
+}
+
+func (d *qdrantProvider) QueryEmbedding(
+	emb []float64,
+	ctx wrapper.HttpContext,
+	log wrapper.Log,
+	callback func(results []QueryResult, ctx wrapper.HttpContext, log wrapper.Log, err error)) error {
+	// 最少需要填写的参数为 vector 和 limit. with_payload 可选，为了直接得到问题答案，所以这里需要
+	// 下面是一个例子
+	// {
+	// 	"vector": [
+	// 	  0.2,
+	// 	  0.1,
+	// 	  0.9,
+	// 	  0.7
+	// 	],
+	// 	"limit": 1
+	// }
+	requestBody, err := json.Marshal(qdrantQueryRequest{
+		Vector:      emb,
+		Limit:       d.config.topK,
+		WithPayload: true,
+	})
+	if err != nil {
+		log.Errorf("[Qdrant] Failed to marshal query embedding: %v", err)
+		return err
+	}
+
+	return d.client.Post(
+		fmt.Sprintf("/collections/%s/points/search", d.config.collectionID),
+		[][2]string{
+			{"Content-Type", "application/json"},
+			{"api-key", d.config.apiKey},
+		},
+		requestBody,
+		func(statusCode int, responseHeaders http.Header, responseBody []byte) {
+			log.Debugf("[Qdrant] Query embedding response: %d, %s", statusCode, responseBody)
+			results, err := d.parseQueryResponse(responseBody, log)
+			if err != nil {
+				err = fmt.Errorf("[Qdrant] Failed to parse query response: %v", err)
+			}
+			callback(results, ctx, log, err)
+		},
+		d.config.timeout,
+	)
+}
+
+func (d *qdrantProvider) parseQueryResponse(responseBody []byte, log wrapper.Log) ([]QueryResult, error) {
+	// 返回的内容例子如下
+	// {
+	// 	"time": 0.002,
+	// 	"status": "ok",
+	// 	"result": [
+	// 	  {
+	// 		"id": 42,
+	// 		"version": 3,
+	// 		"score": 0.75,
+	// 		"payload": {
+	// 		  "question": "London",
+	// 		  "answer": "green"
+	// 		},
+	// 		"shard_key": "region_1",
+	// 		"order_value": 42
+	// 	  }
+	// 	]
+	// }
+	if !gjson.GetBytes(responseBody, "result.0.score").Exists() {
+		log.Errorf("[Qdrant] No distance found in response body: %s", responseBody)
+		return nil, errors.New("[Qdrant] No distance found in response body")
+	}
+
+	if !gjson.GetBytes(responseBody, "result.0.payload.answer").Exists() {
+		log.Errorf("[Qdrant] No answer found in response body: %s", responseBody)
+		return nil, errors.New("[Qdrant] No answer found in response body")
+	}
+
+	resultNum := gjson.GetBytes(responseBody, "result.#").Int()
+	results := make([]QueryResult, 0, resultNum)
+	for i := 0; i < int(resultNum); i++ {
+		result := QueryResult{
+			Text:   gjson.GetBytes(responseBody, fmt.Sprintf("result.%d.payload.question", i)).String(),
+			Score:  gjson.GetBytes(responseBody, fmt.Sprintf("result.%d.score", i)).Float(),
+			Answer: gjson.GetBytes(responseBody, fmt.Sprintf("result.%d.payload.answer", i)).String(),
+		}
+		results = append(results, result)
+	}
+
+	return results, nil
+}
--- a/plugins/wasm-go/extensions/ai-cache/vector/weaviate.go
+++ b/plugins/wasm-go/extensions/ai-cache/vector/weaviate.go
@@ -0,0 +1,188 @@
+package vector
+
+import (
+	"encoding/json"
+	"errors"
+	"fmt"
+	"net/http"
+
+	"github.com/alibaba/higress/plugins/wasm-go/pkg/wrapper"
+	"github.com/tidwall/gjson"
+)
+
+type weaviateProviderInitializer struct{}
+
+func (c *weaviateProviderInitializer) ValidateConfig(config ProviderConfig) error {
+	if len(config.collectionID) == 0 {
+		return errors.New("[Weaviate] collectionID is required")
+	}
+	if len(config.serviceName) == 0 {
+		return errors.New("[Weaviate] serviceName is required")
+	}
+	return nil
+}
+
+func (c *weaviateProviderInitializer) CreateProvider(config ProviderConfig) (Provider, error) {
+	return &WeaviateProvider{
+		config: config,
+		client: wrapper.NewClusterClient(wrapper.FQDNCluster{
+			FQDN: config.serviceName,
+			Host: config.serviceHost,
+			Port: int64(config.servicePort),
+		}),
+	}, nil
+}
+
+type WeaviateProvider struct {
+	config ProviderConfig
+	client wrapper.HttpClient
+}
+
+func (c *WeaviateProvider) GetProviderType() string {
+	return PROVIDER_TYPE_WEAVIATE
+}
+
+func (d *WeaviateProvider) QueryEmbedding(
+	emb []float64,
+	ctx wrapper.HttpContext,
+	log wrapper.Log,
+	callback func(results []QueryResult, ctx wrapper.HttpContext, log wrapper.Log, err error)) error {
+	// 最少需要填写的参数为 class, vector
+	// 下面是一个例子
+	// {"query": "{ Get { Higress ( limit: 2 nearVector: { vector: [0.1, 0.2, 0.3] } ) { question _additional { distance } } } }"}
+	embString, err := json.Marshal(emb)
+	if err != nil {
+		log.Errorf("[Weaviate] Failed to marshal query embedding: %v", err)
+		return err
+	}
+	// 这里默认按照 distance 进行升序，所以不用再次排序
+	graphql := fmt.Sprintf(`
+	{
+	  Get {
+	    %s (
+	      limit: %d
+	      nearVector: {
+	        vector: %s
+	      }
+	    ) {
+		  question
+		  answer
+	      _additional {
+	        distance
+	      }
+	    }
+	  }
+	}
+	`, d.config.collectionID, d.config.topK, embString)
+
+	requestBody, err := json.Marshal(weaviateQueryRequest{
+		Query: graphql,
+	})
+
+	if err != nil {
+		log.Errorf("[Weaviate] Failed to marshal query embedding request body: %v", err)
+		return err
+	}
+
+	err = d.client.Post(
+		"/v1/graphql",
+		[][2]string{
+			{"Content-Type", "application/json"},
+			{"Authorization", fmt.Sprintf("Bearer %s", d.config.apiKey)},
+		},
+		requestBody,
+		func(statusCode int, responseHeaders http.Header, responseBody []byte) {
+			log.Debugf("[Weaviate] Query embedding response: %d, %s", statusCode, responseBody)
+			results, err := d.parseQueryResponse(responseBody, log)
+			if err != nil {
+				err = fmt.Errorf("[Weaviate] Failed to parse query response: %v", err)
+			}
+			callback(results, ctx, log, err)
+		},
+		d.config.timeout,
+	)
+	return err
+}
+
+func (d *WeaviateProvider) UploadAnswerAndEmbedding(
+	queryString string,
+	queryEmb []float64,
+	queryAnswer string,
+	ctx wrapper.HttpContext,
+	log wrapper.Log,
+	callback func(ctx wrapper.HttpContext, log wrapper.Log, err error)) error {
+	// 最少需要填写的参数为 class, vector 和 question 和 answer
+	// 下面是一个例子
+	// {"class": "Higress", "vector": [0.1, 0.2, 0.3], "properties": {"question": "这里是问题", "answer": "这里是答案"}}
+	requestBody, err := json.Marshal(weaviateInsertRequest{
+		Class:      d.config.collectionID,
+		Vector:     queryEmb,
+		Properties: weaviateProperties{Question: queryString, Answer: queryAnswer}, // queryString 指的是用户查询的问题
+	})
+
+	if err != nil {
+		log.Errorf("[Weaviate] Failed to marshal upload embedding request body: %v", err)
+		return err
+	}
+
+	return d.client.Post(
+		"/v1/objects",
+		[][2]string{
+			{"Content-Type", "application/json"},
+			{"Authorization", fmt.Sprintf("Bearer %s", d.config.apiKey)},
+		},
+		requestBody,
+		func(statusCode int, responseHeaders http.Header, responseBody []byte) {
+			log.Debugf("[Weaviate] statusCode: %d, responseBody: %s", statusCode, string(responseBody))
+			callback(ctx, log, err)
+		},
+		d.config.timeout,
+	)
+}
+
+type weaviateProperties struct {
+	Question string `json:"question"`
+	Answer   string `json:"answer"`
+}
+
+type weaviateInsertRequest struct {
+	Class      string             `json:"class"`
+	Vector     []float64          `json:"vector"`
+	Properties weaviateProperties `json:"properties"`
+}
+
+type weaviateQueryRequest struct {
+	Query string `json:"query"`
+}
+
+func (d *WeaviateProvider) parseQueryResponse(responseBody []byte, log wrapper.Log) ([]QueryResult, error) {
+	log.Infof("[Weaviate] queryResp: %s", string(responseBody))
+
+	if !gjson.GetBytes(responseBody, fmt.Sprintf("data.Get.%s.0._additional.distance", d.config.collectionID)).Exists() {
+		log.Errorf("[Weaviate] No distance found in response body: %s", responseBody)
+		return nil, errors.New("[Weaviate] No distance found in response body")
+	}
+
+	if !gjson.GetBytes(responseBody, fmt.Sprintf("data.Get.%s.0.question", d.config.collectionID)).Exists() {
+		log.Errorf("[Weaviate] No question found in response body: %s", responseBody)
+		return nil, errors.New("[Weaviate] No question found in response body")
+	}
+
+	if !gjson.GetBytes(responseBody, fmt.Sprintf("data.Get.%s.0.answer", d.config.collectionID)).Exists() {
+		log.Errorf("[Weaviate] No answer found in response body: %s", responseBody)
+		return nil, errors.New("[Weaviate] No answer found in response body")
+	}
+
+	resultNum := gjson.GetBytes(responseBody, fmt.Sprintf("data.Get.%s.#", d.config.collectionID)).Int()
+	results := make([]QueryResult, 0, resultNum)
+	for i := 0; i < int(resultNum); i++ {
+		result := QueryResult{
+			Text:   gjson.GetBytes(responseBody, fmt.Sprintf("data.Get.%s.%d.question", d.config.collectionID, i)).String(),
+			Score:  gjson.GetBytes(responseBody, fmt.Sprintf("data.Get.%s.%d._additional.distance", d.config.collectionID, i)).Float(),
+			Answer: gjson.GetBytes(responseBody, fmt.Sprintf("data.Get.%s.%d.answer", d.config.collectionID, i)).String(),
+		}
+		results = append(results, result)
+	}
+
+	return results, nil
+}
--- a/plugins/wasm-go/extensions/ai-history/go.mod
+++ b/plugins/wasm-go/extensions/ai-history/go.mod
@@ -8,8 +8,8 @@ replace github.com/alibaba/higress/plugins/wasm-go => ../..

 require (
 	github.com/alibaba/higress/plugins/wasm-go v1.3.6-0.20240528060522-53bccf89f441
-	github.com/higress-group/proxy-wasm-go-sdk v0.0.0-20240711023527-ba358c48772f
-	github.com/tidwall/gjson v1.14.3
+	github.com/higress-group/proxy-wasm-go-sdk v1.0.0
+	github.com/tidwall/gjson v1.17.3
 	github.com/tidwall/resp v0.1.1
 )

--- a/Show More
+++ b/Show More
Author	SHA1	Message	Date
zty98751	32f9a5ff32	fix istio commit	2025-01-15 15:29:44 +08:00
澄潭	6f95297b80	Release 2.0.6-rc.2 (#1671 )	2025-01-14 20:10:53 +08:00
Kent Dong	95426d5ccf	fix: Fix a typo in the README files of ai-statistics plugin (#1670 )	2025-01-14 13:39:55 +08:00
澄潭	a05b6b1e9d	add ai_log field (#1669 )	2025-01-14 10:03:24 +08:00
Jun	d0628344da	add higress architecture doc (#1662 )	2025-01-14 09:48:32 +08:00
韩贤涛	a1bf315b13	fix: resolve blocking issue with minimax responses in ai-proxy (#1663 )	2025-01-14 09:43:19 +08:00
mamba	b3d9123d59	[frontend-gray] 微前端灰度场景,支持 IncludePathPrefixes字段 (#1666 )	2025-01-13 16:24:51 +08:00
rinfx	817061c6cc	remove dependency for ai-statistic (#1660 )	2025-01-10 13:43:29 +08:00
rinfx	ea0d5e7564	Improve ai plugins (#1657 ) Co-authored-by: Kent Dong <ch3cho@qq.com>	2025-01-09 22:04:51 +08:00
澄潭	2a89c3bb70	Optimize wasmplugin proto (#1656 )	2025-01-09 13:19:46 +08:00
johnlanni	a570c72504	Update Chart.lock	2025-01-08 17:14:27 +08:00
澄潭	ab1316dfe1	rel: Release 2.0.6-rc.1 (#1653 )	2025-01-08 17:08:09 +08:00
澄潭	e97448b71b	Update metrics & enable lds cache (#1650 )	2025-01-08 16:49:23 +08:00
澄潭	6820a06a99	fix tls version annotation (#1652 )	2025-01-08 15:31:39 +08:00
澄潭	4733af849d	Update README.md	2025-01-08 11:30:29 +08:00
yunmaoQu	1c2330e33b	feat: add TLS version annotation support for per-rule configuration (#1592 ) Co-authored-by: 澄潭 <zty98751@alibaba-inc.com>	2025-01-06 21:29:09 +08:00
澄潭	61fef0ecf8	Release 2.0.5 (#1646 )	2025-01-06 19:42:18 +08:00
澄潭	d29b8d7ca8	fix ai proxy checkStream (#1645 )	2025-01-06 15:30:02 +08:00
澄潭	2501895b66	ai-cache update body buffer limit size (#1644 )	2025-01-06 14:53:29 +08:00
Kent Dong	187a7b5408	fix: Enlarge the default retry timeout in ai-proxy (#1640 )	2025-01-03 11:19:40 +08:00
Jingze	00be491d02	feat: support github provider for oidc wasm plugin (#1639 )	2025-01-02 10:01:54 +08:00
ayanami-desu	2d74c48e8a	Add cohere embedding for ai-cache (#1572 )	2024-12-27 17:48:44 +08:00
澄潭	6dc4d43df5	optimize ai cache (#1626 )	2024-12-27 10:10:57 +08:00
rinfx	2a4e55d46f	move oidcHandler from global to pluginconfig (#1601 )	2024-12-26 19:15:20 +08:00
Se7en	579c986915	feat: retry failed request (#1590 )	2024-12-26 18:30:50 +08:00
Kent Dong	380717ae3d	fix: Make opa listen to all IPs (#1621 )	2024-12-26 17:41:28 +08:00
Kent Dong	8f3723f554	feat: Support setting gateway.unprivilegedPortSupported manually (#1616 )	2024-12-23 19:45:47 +08:00
VinciWu557	909cc0f088	feat: AI 代理 Wasm 插件接入 Together AI (#1617 )	2024-12-23 15:39:56 +08:00
007gzs	4eaf204737	Enhance the capabilities of the AI Intent plugin (#1605 )	2024-12-20 10:25:17 +08:00
澄潭	748bcb083a	redis wrapper support lazy init and database options (#1602 )	2024-12-19 16:22:56 +08:00
澄潭	39c007d045	optimize ai proxy (#1603 )	2024-12-19 16:22:35 +08:00
rinfx	d74d327b68	bugfix: cannot parse content if one streaming body has multi chunks (#1606 )	2024-12-19 16:21:57 +08:00
澄潭	be27726721	Update CODEOWNERS	2024-12-19 14:36:11 +08:00
澄潭	34cc1c0632	Update README.md	2024-12-18 17:02:28 +08:00
澄潭	5694475872	Update README.md	2024-12-18 16:59:03 +08:00
rinfx	2f5709a93e	qwen bailian compatible bug fix (#1597 )	2024-12-17 16:57:31 +08:00
StarryNight	2a200cdd42	AI proxy return unified status in header phase (#1588 )	2024-12-16 18:41:38 +08:00
rinfx	ec39d56731	AI observability upgrade (#1587 ) Co-authored-by: Kent Dong <ch3cho@qq.com>	2024-12-16 10:27:49 +08:00
韩贤涛	8544fa604d	feat: support choosing chatCompletionV2 or chatCompletionPro API for minimax provider (#1593 )	2024-12-15 15:12:00 +08:00
mirror	0ba63e5dd4	fix: default port of static service in ai-cache plugin (#1591 )	2024-12-13 19:03:26 +08:00
mirror	441408c593	docs: fix typos in ai-quota document (#1589 )	2024-12-13 08:56:43 +08:00
duxin40	be57960c22	Support OpenAI embedding. (#1542 )	2024-12-11 11:42:51 +08:00
rinfx	f32020068a	bugfix and extend ai log (#1576 )	2024-12-09 20:39:13 +08:00
澄潭	1a8fce48f0	Update release-hgctl.yaml	2024-12-06 14:01:18 +08:00
澄潭	85c7b1f501	rel: Release 2.0.4 (#1571 )	2024-12-06 13:52:03 +08:00
pepesi	8f660211e3	feat: ai-proxy support custom error handler by cover util.ErrorHandler (#1537 )	2024-12-06 11:47:50 +08:00
rinfx	433227323d	extension mechanism for custom logs and span attributes (#1451 )	2024-12-05 18:39:00 +08:00
pepesi	b36e5ea26b	feat: allow cover api-version when use ai-proxy azure provider (#1535 )	2024-12-05 13:41:02 +08:00
rinfx	ce66ff68ce	solve aliyun lvwang content length limit problem (#1569 )	2024-12-05 13:39:20 +08:00
pepesi	d026f0fca5	feat: ai-proxy support dashscope-finance (#1554 )	2024-12-05 11:48:09 +08:00
rinfx	22790aa149	fix moonshot usage compatible problem (#1568 )	2024-12-05 11:35:25 +08:00
澄潭	7ce6d7aba1	fix xds cache (#1559 )	2024-12-04 00:55:29 +08:00
Se7en	e705a0344f	fix: qwen stream issue (#1564 )	2024-12-03 13:10:47 +08:00
澄潭	d6094974c2	update ai proxy go mod (#1556 )	2024-12-02 14:41:55 +08:00
mamba	6187be97e5	fix: 🐛 frontend-grayurl 解析不正确导致路由失败 (#1550 )	2024-11-29 13:09:05 +08:00
澄潭	bb64b43f23	set concurrency argument of proxy by cpu limit/request (#1552 )	2024-11-28 16:55:57 +08:00
澄潭	ca7458cf1c	Optimize the overall log output (#1549 )	2024-11-27 20:44:34 +08:00
Se7en	ee2dd76ae1	feat: migrate baidu provider to v2 api (#1527 )	2024-11-27 20:12:00 +08:00
pepesi	8154cf95f1	feat: support custom log (#1521 )	2024-11-27 20:11:29 +08:00
澄潭	a7593381e1	fix ai fallback (#1541 )	2024-11-25 16:48:59 +08:00
澄潭	e68a8ac25f	add model-mapper plugin & optimize model-router plugin (#1538 )	2024-11-22 22:24:42 +08:00
Kent Dong	96575b982e	fix: Refresh go.mod and go.sum file contents (#1525 )	2024-11-22 13:34:55 +08:00
EnableAsync	c2d405b2a7	feat: Enhance ai-cache Plugin with Vector Similarity-Based LLM Cache Recall and Multi-DB Support (#1248 )	2024-11-21 16:57:41 +08:00
Jingze	6efb3109f2	fix: update oidc plugin go.mod dependencies (#1522 )	2024-11-19 17:33:42 +08:00
Se7en	1b1c08afb7	fix: apitoken failover for coze (#1515 )	2024-11-18 15:36:26 +08:00
Se7en	d24123a55f	feat: implement apiToken failover mechanism (#1256 )	2024-11-16 19:03:09 +08:00
澄潭	f2a5df3949	use the body returned by the ext auth server when auth fails (#1510 )	2024-11-14 18:50:33 +08:00
澄潭	ebc5b2987e	fix compile of wasm cpp plugins (#1511 )	2024-11-14 18:49:21 +08:00
007gzs	ca97cbd75a	fix workflows build-and-push-wasm-plugin-image (#1508 )	2024-11-13 17:39:24 +08:00
hanans426	a787e237ce	增加快速部署到阿里云的部署方案 (#1506 )	2024-11-13 16:26:55 +08:00
纪卓志	6a1bf90d42	feat: supports custom prepare build script (#1490 )	2024-11-12 13:45:28 +08:00
007gzs	60e476da87	fix example sse build error (#1503 )	2024-11-11 17:47:26 +08:00
rinfx	2cb8558cda	Optimize AI security guard plugin (#1473 ) Co-authored-by: Kent Dong <ch3cho@qq.com>	2024-11-11 14:49:17 +08:00
littlejian	4d1a037942	feat: Automatically generating markdown documentation for helm charts with helm-docs (#1496 )	2024-11-11 11:34:38 +08:00
xingyunyang01	39b6eac9d0	AI Agent plugin adds JSON formatting output feature (#1374 )	2024-11-11 11:11:02 +08:00