封装fastAPI openAI接口规范

2025-11-04 23:57:37 +08:00
parent a75989c3e6
commit 8a9aef87aa
15 changed files with 904 additions and 7 deletions
--- a/.DS_Store
+++ b/.DS_Store
--- a/.dockerignore
+++ b/.dockerignore
@@ -0,0 +1,45 @@
+# Git
+.git
+.gitignore
+
+# Python
+__pycache__/
+*.pyc
+*.pyo
+*.pyd
+.Python
+env/
+pip-log.txt
+pip-delete-this-directory.txt
+.tox/
+.coverage
+.coverage.*
+.cache
+nosetests.xml
+coverage.xml
+*.cover
+*.log
+.git
+.mypy_cache
+.pytest_cache
+.hypothesis
+
+# IDE
+.vscode/
+.idea/
+*.swp
+*.swo
+*~
+
+# OS
+.DS_Store
+.DS_Store?
+._*
+.Spotlight-V100
+.Trashes
+ehthumbs.db
+Thumbs.db
+
+# Project specific
+*.md
+!README.md
--- a/31
+++ b/31
@@ -0,0 +1,31 @@
+# 使用Python 3.10作为基础镜像
+FROM python:3.12-slim
+
+# 设置工作目录
+WORKDIR /app
+
+# 设置环境变量
+ENV PYTHONPATH=/app
+ENV PYTHONUNBUFFERED=1
+
+# 安装系统依赖
+RUN apt-get update && apt-get install -y \
+    gcc \
+    g++ \
+    && rm -rf /var/lib/apt/lists/*
+
+# 复制项目文件
+COPY pyproject.toml ./
+COPY fastapi_server/requirements.txt ./fastapi_server/
+COPY lang_agent/ ./lang_agent/
+COPY fastapi_server/ ./fastapi_server/
+
+# 安装Python依赖
+RUN pip install --no-cache-dir -r fastapi_server/requirements.txt
+RUN pip install --no-cache-dir -e .
+
+# 暴露端口
+EXPOSE 8488
+
+# 启动命令
+CMD ["python", "fastapi_server/server.py"]
--- a/docker-compose.yml
+++ b/docker-compose.yml
@@ -0,0 +1,25 @@
+version: '3.8'
+
+services:
+  lang-agent-api:
+    build: .
+    container_name: lang-agent-api
+    ports:
+      - "8488:8488"
+    env_file:
+      - ./.env
+    environment:
+      - PYTHONPATH=/app
+      - PYTHONUNBUFFERED=1
+      - RAG_FOLDER_PATH=/app/assets/xiaozhan_emb
+    volumes:
+      - ./configs:/app/configs
+      - ./scripts:/app/scripts
+      - ./assets:/app/assets
+    restart: unless-stopped
+    healthcheck:
+      test: ["CMD", "python", "-c", "import urllib.request; urllib.request.urlopen('http://localhost:8488/health')"]
+      interval: 30s
+      timeout: 10s
+      retries: 3
+      start_period: 40s
--- a/fastapi_server/Dockerfile.api
+++ b/fastapi_server/Dockerfile.api
@@ -0,0 +1,20 @@
+# 使用Python 3.9作为基础镜像
+FROM python:3.9-slim
+
+# 设置工作目录
+WORKDIR /app
+
+# 复制requirements文件
+COPY requirements.txt .
+
+# 安装Python依赖
+RUN pip install --no-cache-dir -r requirements.txt
+
+# 复制项目文件
+COPY . .
+
+# 暴露端口
+EXPOSE 8488
+
+# 启动命令
+CMD ["python", "server.py"]
--- a/fastapi_server/OpenAI_API_README.md
+++ b/fastapi_server/OpenAI_API_README.md
@@ -0,0 +1,220 @@
+# Lang Agent OpenAI 兼容API
+
+这是一个符合OpenAI接口规范的聊天API，允许用户使用与OpenAI API相同的方式访问您的Lang Agent服务。
+
+## 快速开始
+
+### 1. 启动服务器
+
+```bash
+cd /path/to/lang-agent/fastapi_server
+python server.py
+```
+
+服务器将在 `http://localhost:8488` 上启动。
+
+### 2. 使用API
+
+#### 使用curl命令
+
+```bash
+curl -X POST "http://localhost:8488/v1/chat/completions" \
+  -H "Authorization: Bearer 123tangledup-ai" \
+  -H "Content-Type: application/json" \
+  -d '{
+      "model": "qwen-plus",
+      "messages": [
+          {
+              "role": "system",
+              "content": "You are a helpful assistant."
+          },
+          {
+              "role": "user",
+              "content": "你是谁？"
+          }
+      ]
+  }'
+```
+
+#### 使用Python requests
+
+```python
+import requests
+
+API_BASE_URL = "http://localhost:8488"
+API_KEY = "123tangledup-ai"
+
+headers = {
+    "Authorization": f"Bearer {API_KEY}",
+    "Content-Type": "application/json"
+}
+
+data = {
+    "model": "qwen-plus",
+    "messages": [
+        {
+            "role": "system",
+            "content": "You are a helpful assistant."
+        },
+        {
+            "role": "user",
+            "content": "你是谁？"
+        }
+    ]
+}
+
+response = requests.post(f"{API_BASE_URL}/v1/chat/completions", headers=headers, json=data)
+print(response.json())
+```
+
+#### 使用OpenAI Python库
+
+```python
+from openai import OpenAI
+
+client = OpenAI(
+    api_key="123tangledup-ai",
+    base_url="http://localhost:8488/v1"
+)
+
+response = client.chat.completions.create(
+    model="qwen-plus",
+    messages=[
+        {"role": "system", "content": "You are a helpful assistant."},
+        {"role": "user", "content": "你是谁？"}
+    ]
+)
+
+print(response.choices[0].message.content)
+```
+
+## API 端点
+
+### 1. 聊天完成 `/v1/chat/completions`
+
+与OpenAI的chat completions API完全兼容。
+
+**请求参数:**
+
+| 参数 | 类型 | 必需 | 默认值 | 描述 |
+|------|------|------|--------|------|
+| model | string | 是 | - | 模型名称 |
+| messages | array | 是 | - | 消息列表 |
+| temperature | number | 否 | 0.7 | 采样温度 |
+| max_tokens | integer | 否 | 500 | 最大生成token数 |
+| stream | boolean | 否 | false | 是否流式返回 |
+| thread_id | integer | 否 | 3 | 线程ID，用于多轮对话 |
+
+**响应格式:**
+
+```json
+{
+  "id": "chatcmpl-abc123",
+  "object": "chat.completion",
+  "created": 1677652288,
+  "model": "qwen-plus",
+  "choices": [
+    {
+      "index": 0,
+      "message": {
+        "role": "assistant",
+        "content": "您好！我是一个AI助手..."
+      },
+      "finish_reason": "stop"
+    }
+  ],
+  "usage": {
+    "prompt_tokens": 56,
+    "completion_tokens": 31,
+    "total_tokens": 87
+  }
+}
+```
+
+### 2. 健康检查 `/health`
+
+检查API服务状态。
+
+**请求:**
+```bash
+GET /health
+```
+
+**响应:**
+```json
+{
+  "status": "healthy"
+}
+```
+
+### 3. API信息 `/`
+
+获取API基本信息。
+
+**请求:**
+```bash
+GET /
+```
+
+**响应:**
+```json
+{
+  "message": "Lang Agent Chat API",
+  "version": "1.0.0",
+  "description": "使用OpenAI格式调用pipeline.invoke的聊天API",
+  "authentication": "Bearer Token (API Key)",
+  "endpoints": {
+    "/v1/chat/completions": "POST - 聊天完成接口，兼容OpenAI格式，需要API密钥验证",
+    "/": "GET - API信息",
+    "/health": "GET - 健康检查接口"
+  }
+}
+```
+
+## 认证
+
+API使用Bearer Token认证。默认API密钥为 `123tangledup-ai`。
+
+在请求头中包含：
+```
+Authorization: Bearer 123tangledup-ai
+```
+
+## 测试脚本
+
+项目提供了两个测试脚本：
+
+1. **Bash脚本** (`test_openai_api.sh`) - 使用curl命令测试API
+2. **Python脚本** (`test_openai_api.py`) - 使用Python requests库测试API
+
+运行测试脚本：
+
+```bash
+# 运行Bash测试脚本
+chmod +x test_openai_api.sh
+./test_openai_api.sh
+
+# 运行Python测试脚本
+python test_openai_api.py
+```
+
+## 与OpenAI API的兼容性
+
+此API完全兼容OpenAI的chat completions API，您可以：
+
+1. 使用任何支持OpenAI API的客户端库
+2. 将base_url更改为`http://localhost:8488/v1`
+3. 使用提供的API密钥进行认证
+
+## 注意事项
+
+1. 确保服务器正在运行且可访问
+2. 流式响应(stream=true)目前可能不完全支持
+3. 模型参数(model)主要用于标识，实际使用的模型由服务器配置决定
+4. 多轮对话使用thread_id参数来维护上下文
+
+## 故障排除
+
+1. **连接错误**: 确保服务器正在运行，检查URL和端口是否正确
+2. **认证错误**: 检查API密钥是否正确设置
+3. **请求格式错误**: 确保请求体是有效的JSON格式，包含所有必需字段
--- a/fastapi_server/README.md
+++ b/fastapi_server/README.md
@@ -0,0 +1,179 @@
+# Lang Agent Chat API
+
+这是一个基于FastAPI的聊天API服务，使用OpenAI格式的请求来调用pipeline.invoke方法进行聊天。
+
+## 功能特点
+
+- 兼容OpenAI API格式的聊天接口
+- 支持多轮对话（通过thread_id）
+- 使用qwen-flash模型
+- 支持流式和非流式响应
+- 提供健康检查接口
+
+## 安装依赖
+
+```bash
+pip install -r requirements.txt
+```
+
+## 环境变量
+
+确保设置以下环境变量：
+
+```bash
+export ALI_API_KEY="your_ali_api_key"
+```
+
+## 运行服务
+
+### 方法1：使用启动脚本
+
+```bash
+./start_server.sh
+```
+
+### 方法2：直接运行Python文件
+
+```bash
+python server.py
+```
+
+服务将在 `http://localhost:8000` 启动。
+
+## API接口
+
+### 聊天完成接口
+
+**端点**: `POST /v1/chat/completions`
+
+**请求格式**:
+```json
+{
+  "model": "qwen-flash",
+  "messages": [
+    {
+      "role": "system",
+      "content": "你是一个有用的助手。"
+    },
+    {
+      "role": "user",
+      "content": "你好，请介绍一下你自己。"
+    }
+  ],
+  "temperature": 0.7,
+  "max_tokens": 1000,
+  "stream": false,
+  "thread_id": 3
+}
+```
+
+**响应格式**:
+```json
+{
+  "id": "chatcmpl-abc123",
+  "object": "chat.completion",
+  "created": 1677652288,
+  "model": "qwen-flash",
+  "choices": [
+    {
+      "index": 0,
+      "message": {
+        "role": "assistant",
+        "content": "你好！我是小盏，是半盏青年茶馆的智能助手..."
+      },
+      "finish_reason": "stop"
+    }
+  ]
+}
+```
+
+### API信息接口
+
+**端点**: `GET /`
+
+返回API的基本信息。
+
+### 健康检查接口
+
+**端点**: `GET /health`
+
+返回服务的健康状态。
+
+## 使用示例
+
+### 使用OpenAI Python客户端库
+
+首先安装OpenAI库：
+
+```bash
+pip install openai
+```
+
+然后使用以下代码：
+
+```python
+from openai import OpenAI
+
+# 设置API基础URL和API密钥（这里使用一个虚拟的密钥，因为我们没有实现认证）
+client = OpenAI(
+    api_key="your-api-key",  # 这里可以使用任意值，因为我们的API没有实现认证
+    base_url="http://localhost:8000/v1"
+)
+
+# 发送聊天请求
+response = client.chat.completions.create(
+    model="qwen-flash",
+    messages=[
+        {"role": "system", "content": "你是一个有用的助手。"},
+        {"role": "user", "content": "你好，请介绍一下你自己。"}
+    ],
+    temperature=0.7,
+    thread_id=1  # 用于多轮对话
+)
+
+print(response.choices[0].message.content)
+```
+
+### 使用curl
+
+```bash
+curl -X POST "http://localhost:8000/v1/chat/completions" \
+-H "Content-Type: application/json" \
+-d '{
+  "model": "qwen-flash",
+  "messages": [
+    {
+      "role": "user",
+      "content": "你好，请介绍一下你自己。"
+    }
+  ]
+}'
+```
+
+### 使用Python requests
+
+```python
+import requests
+
+url = "http://localhost:8000/v1/chat/completions"
+headers = {"Content-Type": "application/json"}
+data = {
+    "model": "qwen-flash",
+    "messages": [
+        {
+            "role": "user",
+            "content": "你好，请介绍一下你自己。"
+        }
+    ]
+}
+
+response = requests.post(url, headers=headers, json=data)
+print(response.json())
+```
+
+## 注意事项
+
+1. 确保已设置正确的API密钥环境变量
+2. 默认使用qwen-flash模型，可以通过修改代码中的配置来更改模型
+3. thread_id用于多轮对话，相同的thread_id会保持对话上下文
+4. 目前stream参数设置为true时，仍会返回非流式响应（可根据需要进一步实现）
--- a/fastapi_server/docker-compose.api.yml
+++ b/fastapi_server/docker-compose.api.yml
@@ -0,0 +1,18 @@
+version: '3.8'
+
+services:
+  lang-agent-api:
+    build:
+      context: .
+      dockerfile: Dockerfile.api
+    ports:
+      - "8488:8488"
+    environment:
+      - PYTHONUNBUFFERED=1
+    restart: unless-stopped
+    healthcheck:
+      test: ["CMD", "curl", "-f", "http://localhost:8488/health"]
+      interval: 30s
+      timeout: 10s
+      retries: 3
+      start_period: 40s
--- a/fastapi_server/openai_client_example.py
+++ b/fastapi_server/openai_client_example.py
@@ -0,0 +1,129 @@
+#!/usr/bin/env python3
+"""
+使用OpenAI Python客户端库调用我们的FastAPI聊天API的示例
+"""
+
+from openai import OpenAI
+import os
+
+# 设置API基础URL和API密钥（这里使用一个虚拟的密钥，因为我们没有实现认证）
+client = OpenAI(
+    api_key="your-api-key",  # 这里可以使用任意值，因为我们的API没有实现认证
+    base_url="http://localhost:8000/v1"
+)
+
+def simple_chat():
+    """简单的聊天示例"""
+    print("=" * 50)
+    print("简单聊天示例")
+    print("=" * 50)
+    
+    response = client.chat.completions.create(
+        model="qwen-flash",
+        messages=[
+            {"role": "user", "content": "你好，请介绍一下你自己。"}
+        ],
+        temperature=0.7,
+        thread_id=1
+    )
+    
+    print(f"助手回复: {response.choices[0].message.content}")
+    print("\n")
+
+def multi_turn_chat():
+    """多轮对话示例"""
+    print("=" * 50)
+    print("多轮对话示例")
+    print("=" * 50)
+    
+    # 第一轮对话
+    print("第一轮对话:")
+    response1 = client.chat.completions.create(
+        model="qwen-flash",
+        messages=[
+            {"role": "user", "content": "你推荐什么茶？"}
+        ],
+        temperature=0.7,
+        thread_id=2
+    )
+    
+    print(f"用户: 你推荐什么茶？")
+    print(f"助手: {response1.choices[0].message.content}")
+    
+    # 第二轮对话，使用相同的thread_id
+    print("\n第二轮对话:")
+    response2 = client.chat.completions.create(
+        model="qwen-flash",
+        messages=[
+            {"role": "user", "content": "为什么推荐这个茶？"}
+        ],
+        temperature=0.7,
+        thread_id=2  # 使用相同的thread_id
+    )
+    
+    print(f"用户: 为什么推荐这个茶？")
+    print(f"助手: {response2.choices[0].message.content}")
+    print("\n")
+
+def system_prompt_example():
+    """使用系统提示的示例"""
+    print("=" * 50)
+    print("系统提示示例")
+    print("=" * 50)
+    
+    response = client.chat.completions.create(
+        model="qwen-flash",
+        messages=[
+            {"role": "system", "content": "你是一个专业的茶艺师，用简洁的语言回答问题，不超过50字。"},
+            {"role": "user", "content": "请介绍一下普洱茶。"}
+        ],
+        temperature=0.3,
+        thread_id=3
+    )
+    
+    print(f"用户: 请介绍一下普洱茶。")
+    print(f"助手: {response.choices[0].message.content}")
+    print("\n")
+
+def interactive_chat():
+    """交互式聊天示例"""
+    print("=" * 50)
+    print("交互式聊天 (输入'quit'退出)")
+    print("=" * 50)
+    
+    thread_id = 4  # 为这个会话分配一个固定的thread_id
+    
+    while True:
+        user_input = input("你: ")
+        if user_input.lower() == 'quit':
+            break
+        
+        try:
+            response = client.chat.completions.create(
+                model="qwen-flash",
+                messages=[
+                    {"role": "user", "content": user_input}
+                ],
+                temperature=0.7,
+                thread_id=thread_id
+            )
+            
+            print(f"助手: {response.choices[0].message.content}")
+        except Exception as e:
+            print(f"错误: {str(e)}")
+
+if __name__ == "__main__":
+    print("使用OpenAI客户端库调用FastAPI聊天API示例")
+    print("注意: 确保服务器在 http://localhost:8000 上运行\n")
+    
+    # 简单聊天示例
+    simple_chat()
+    
+    # 多轮对话示例
+    multi_turn_chat()
+    
+    # 系统提示示例
+    system_prompt_example()
+    
+    # 交互式聊天示例
+    interactive_chat()
--- a/fastapi_server/requirements.txt
+++ b/fastapi_server/requirements.txt
@@ -0,0 +1,24 @@
+fastapi>=0.104.0
+uvicorn>=0.24.0
+pydantic>=2.0.0,<2.12
+loguru>=0.7.0
+python-dotenv>=1.0.0
+langchain==1.0
+langchain-core>=0.1.0
+langchain-community
+langchain-openai
+langchain-mcp-adapters
+langgraph>=0.0.40
+tyro>=0.7.0
+commentjson>=0.9.0
+matplotlib>=3.7.0
+Pillow>=10.0.0
+jax>=0.4.0
+httpx[socks]
+dashscope
+websockets>=11.0.3
+mcp>=1.8.1
+mcp-proxy>=0.8.2
+faiss-cpu
+fastmcp
+pandas
--- a/fastapi_server/server.py
+++ b/fastapi_server/server.py
@@ -0,0 +1,163 @@
+from fastapi import FastAPI, HTTPException, Depends, Security
+from fastapi.middleware.cors import CORSMiddleware
+from fastapi.security import HTTPBearer, HTTPAuthorizationCredentials
+from pydantic import BaseModel, Field
+from typing import List, Optional, Dict, Any, Union
+import os
+import sys
+import time
+import uvicorn
+from loguru import logger
+
+# 添加父目录到系统路径，以便导入lang_agent模块
+sys.path.append(os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
+
+from lang_agent.pipeline import Pipeline, PipelineConfig
+
+# 定义OpenAI格式的请求模型
+class ChatMessage(BaseModel):
+    role: str = Field(..., description="消息角色，可以是 'system', 'user', 'assistant'")
+    content: str = Field(..., description="消息内容")
+
+class ChatCompletionRequest(BaseModel):
+    model: str = Field(default="qwen-flash", description="模型名称")
+    messages: List[ChatMessage] = Field(..., description="对话消息列表")
+    temperature: Optional[float] = Field(default=0.7, description="采样温度")
+    max_tokens: Optional[int] = Field(default=500, description="最大生成token数")
+    stream: Optional[bool] = Field(default=False, description="是否流式返回")
+    thread_id: Optional[int] = Field(default=3, description="线程ID，用于多轮对话")
+
+class ChatCompletionResponseChoice(BaseModel):
+    index: int
+    message: ChatMessage
+    finish_reason: str
+
+class ChatCompletionResponseUsage(BaseModel):
+    prompt_tokens: int
+    completion_tokens: int
+    total_tokens: int
+
+class ChatCompletionResponse(BaseModel):
+    id: str
+    object: str = "chat.completion"
+    created: int
+    model: str
+    choices: List[ChatCompletionResponseChoice]
+    usage: Optional[ChatCompletionResponseUsage] = None
+
+# 初始化FastAPI应用
+app = FastAPI(title="Lang Agent Chat API", description="使用OpenAI格式调用pipeline.invoke的聊天API")
+
+# 设置API密钥
+API_KEY = "123tangledup-ai"
+
+# 创建安全方案
+security = HTTPBearer()
+
+# 验证API密钥的依赖项
+async def verify_api_key(credentials: HTTPAuthorizationCredentials = Security(security)):
+    if credentials.credentials != API_KEY:
+        raise HTTPException(
+            status_code=401,
+            detail="无效的API密钥",
+            headers={"WWW-Authenticate": "Bearer"},
+        )
+    return credentials
+
+# 添加CORS中间件
+app.add_middleware(
+    CORSMiddleware,
+    allow_origins=["*"],
+    allow_credentials=True,
+    allow_methods=["*"],
+    allow_headers=["*"],
+)
+
+# 初始化Pipeline
+pipeline_config = PipelineConfig()
+pipeline_config.llm_name = "qwen-flash"
+pipeline_config.llm_provider = "openai"
+pipeline_config.base_url = "https://dashscope.aliyuncs.com/compatible-mode/v1"
+
+pipeline = Pipeline(pipeline_config)
+
+@app.post("/v1/chat/completions", response_model=ChatCompletionResponse)
+async def chat_completions(
+    request: ChatCompletionRequest, 
+    credentials: HTTPAuthorizationCredentials = Depends(verify_api_key)
+):
+    """
+    使用OpenAI格式的聊天完成API
+    """
+    try:
+        # 提取用户消息
+        user_message = None
+        system_message = None
+        
+        for message in request.messages:
+            if message.role == "user":
+                user_message = message.content
+            elif message.role == "system":
+                system_message = message.content
+        
+        if not user_message:
+            raise HTTPException(status_code=400, detail="缺少用户消息")
+        
+        # 调用pipeline的chat方法
+        response_content = pipeline.chat(
+            inp=user_message,
+            as_stream=request.stream,
+            thread_id=request.thread_id
+        )
+        
+        # 构建响应
+        response = ChatCompletionResponse(
+            id=f"chatcmpl-{os.urandom(12).hex()}",
+            created=int(time.time()),
+            model=request.model,
+            choices=[
+                ChatCompletionResponseChoice(
+                    index=0,
+                    message=ChatMessage(role="assistant", content=response_content),
+                    finish_reason="stop"
+                )
+            ]
+        )
+        
+        return response
+    
+    except Exception as e:
+        logger.error(f"处理聊天请求时出错: {str(e)}")
+        raise HTTPException(status_code=500, detail=f"内部服务器错误: {str(e)}")
+
+@app.get("/")
+async def root():
+    """
+    根路径，返回API信息
+    """
+    return {
+        "message": "Lang Agent Chat API",
+        "version": "1.0.0",
+        "description": "使用OpenAI格式调用pipeline.invoke的聊天API",
+        "authentication": "Bearer Token (API Key)",
+        "endpoints": {
+            "/v1/chat/completions": "POST - 聊天完成接口，兼容OpenAI格式，需要API密钥验证",
+            "/": "GET - API信息",
+            "/health": "GET - 健康检查接口"
+        }
+    }
+
+@app.get("/health")
+async def health_check():
+    """
+    健康检查接口
+    """
+    return {"status": "healthy"}
+
+if __name__ == "__main__":
+    uvicorn.run(
+        "server:app",
+        host="0.0.0.0",
+        port=8488,
+        reload=True
+    )
--- a/fastapi_server/start_server.sh
+++ b/fastapi_server/start_server.sh
@@ -0,0 +1,19 @@
+#!/bin/bash
+
+echo "启动Lang Agent Chat API服务器..."
+
+# 检查Python环境
+if ! command -v python &> /dev/null; then
+    echo "错误: 未找到Python。请确保Python已安装并添加到PATH中。"
+    exit 1
+fi
+
+# 检查环境变量
+if [ -z "$ALI_API_KEY" ]; then
+    echo "警告: 未设置ALI_API_KEY环境变量。请确保已设置此变量。"
+    echo "例如: export ALI_API_KEY='your_api_key'"
+fi
+
+# 启动服务器
+cd "$(dirname "$0")"
+python server.py
--- a/lang_agent/rag/simple.py
+++ b/lang_agent/rag/simple.py
@@ -21,8 +21,21 @@ class SimpleRagConfig(ToolConfig, KeyConfig):
    model_name:str = "text-embedding-v4"
    """embedding model name"""

-    folder_path:str = "/home/smith/projects/work/langchain-agent/assets/xiaozhan_emb"
-    """path to local database"""
+    folder_path:str = "/Users/jeremygan/Desktop/TangledupAI/lang-agent/assets/xiaozhan_emb"
+    """path to docker database"""
+    
+    # @property
+    # def folder_path(self) -> str:
+    #     """Dynamically determine the folder path for the vector store"""
+    #     # Check if environment variable is set
+    #     env_path = os.environ.get("RAG_FOLDER_PATH")
+    #     if env_path:
+    #         return env_path
+            
+    #     # Default to relative path from current working directory
+    #     return os.path.join(os.getcwd(), "assets", "xiaozhan_emb")
+
+    



@@ -31,8 +44,19 @@ class SimpleRag(LangToolBase):
        self.config = config
        self.emb = QwenEmbeddings(self.config.api_key,
                                  self.config.model_name)
+        
+        # Determine the folder path dynamically
+        # folder_path = os.environ.get("RAG_FOLDER_PATH")
+        # if not folder_path:
+        #     # Default to relative path from current working directory
+        #     folder_path = os.path.join(os.getcwd(), "assets", "xiaozhan_emb")
+        
+        # logger.info(f"Loading FAISS index from: {folder_path}")
+
+        folder_path = "/Users/jeremygan/Desktop/TangledupAI/lang-agent/assets/xiaozhan_emb"
+        
        self.vec_store = FAISS.load_local(
-            folder_path=self.config.folder_path,
+            folder_path=folder_path,
            embeddings=self.emb,
            allow_dangerous_deserialization=True  # Required for LangChain >= 0.1.1
        )
--- a/lang_agent/test.py
+++ b/lang_agent/test.py
--- a/lang_agent/tool_manager.py
+++ b/lang_agent/tool_manager.py
@@ -6,14 +6,14 @@ import inspect
 import asyncio
 import os.path as osp
 from loguru import logger
-from fastmcp.tools.tool import FunctionTool
+from fastmcp.tools.tool import Tool

 from lang_agent.config import InstantiateConfig, ToolConfig
 from lang_agent.base import LangToolBase

 from lang_agent.rag.simple import SimpleRagConfig
 from lang_agent.dummy.calculator import CalculatorConfig
-from catering_end.lang_tool import CartToolConfig, CartTool
+# from catering_end.lang_tool import CartToolConfig, CartTool

 from langchain_core.tools.structured import StructuredTool
 import jax
@@ -26,7 +26,7 @@ class ToolManagerConfig(InstantiateConfig):
    # tool configs here; MUST HAVE 'config' in name and must be dataclass
    rag_config: SimpleRagConfig = field(default_factory=SimpleRagConfig)

-    cart_config: CartToolConfig = field(default_factory=CartToolConfig)
+    # cart_config: CartToolConfig = field(default_factory=CartToolConfig)

    calc_config: CalculatorConfig = field(default_factory=CalculatorConfig)

@@ -78,7 +78,7 @@ class ToolManager:
    def _get_tool_fnc(self, tool_obj:LangToolBase)->List:
        fnc_list = []
        for fnc in tool_obj.get_tool_fnc():
-            if isinstance(fnc, FunctionTool):
+            if isinstance(fnc, Tool):
                fnc = fnc.fn
            fnc_list.append(fnc)