跳至主要内容

HTTP API

Cognitive Modules 提供 HTTP API 服务，方便集成到工作流平台（Coze、n8n、Dify 等）。

安装​

# 全局安装
npm install -g cogn

# 或使用 npx
npx cogn --help

启动服务​

# 启动 API 服务
cog serve --port 8000

# 或指定 host
cog serve --host 0.0.0.0 --port 8000

服务启动后，访问：

API 端点​

方法端点说明
GET/API 信息
GET/health健康检查
GET/modules列出所有模块
GET/modules/{name}获取模块详情
POST/run运行模块

运行模块​

请求​

curl -X POST http://localhost:8000/run \
  -H "Content-Type: application/json" \
  -d '{
    "module": "code-reviewer",
    "args": "def login(u,p): return db.query(f\"SELECT * FROM users WHERE name={u}\")"
  }'

响应​

{
  "ok": true,
  "data": {
    "issues": [
      {
        "type": "security",
        "severity": "critical",
        "description": "SQL 注入漏洞"
      }
    ],
    "confidence": 0.95,
    "rationale": "检测到字符串拼接 SQL"
  },
  "module": "code-reviewer",
  "provider": "openai"
}

指定 LLM 提供商​

curl -X POST http://localhost:8000/run \
  -H "Content-Type: application/json" \
  -d '{
    "module": "code-reviewer",
    "args": "your code",
    "provider": "anthropic",
    "model": "claude-3-5-sonnet-20241022"
  }'

工作流集成示例​

n8n​

  1. 添加 HTTP Request 节点
  2. 配置：
    • Method: POST
    • URL: http://localhost:8000/run
    • Body Content Type: JSON
    • Body: 
      {
        "module": "code-reviewer",
        "args": "{{ $json.code }}"
      }

Coze / Dify​

使用 HTTP 插件或代码块调用 API：

import requests

response = requests.post(
    "http://your-server:8000/run",
    json={
        "module": "code-reviewer",
        "args": input_code
    }
)
result = response.json()

Make / Zapier​

  1. 使用 HTTP 模块
  2. POST 到 /run 端点
  3. 解析 JSON 响应

Docker 部署​

FROM python:3.11-slim

RUN pip install cognitive-modules[server,openai]

ENV OPENAI_API_KEY=your-key

EXPOSE 8000

CMD ["cogn", "serve", "--host", "0.0.0.0", "--port", "8000"]
docker build -t cognitive-api .
docker run -p 8000:8000 -e OPENAI_API_KEY=sk-xxx cognitive-api

环境变量​

服务启动前需要配置 LLM API Key：

export OPENAI_API_KEY=sk-xxx
# 或
export ANTHROPIC_API_KEY=sk-xxx

cogn serve

健康检查​

curl http://localhost:8000/health

响应：

{
  "status": "healthy",
  "version": "1.3.0",
  "providers": {
    "openai": true,
    "anthropic": false,
    "minimax": false,
    "deepseek": false
  }
}