← 全部项目

Weixin-AI-DeepSeek-Bridge

Self-hosted WeChat AI assistant with OpenClaw, DeepSeek, and a full admin console. 基于 OpenClaw 和 DeepSeek 的自托管微信 AI 助手，内置模型、人设、技能、服务和日志管理后台。

JavaScript ★ 0 ⑂ 0 活跃 更新于 2026-07-02
在 GitHub 上查看 →

README

Weixin AI DeepSeek Bridge

This project runs a DeepSeek chat bridge for Tencent Weixin AI / OpenClaw.
It also includes an admin console for model settings, persona files, skills, services, and logs.

Features

  • POST /chat: simple chat API for local tests.

  • POST /v1/chat/completions: OpenAI-compatible chat API.

  • GET /v1/models: returns the current configured model.

  • Supports stream=true.

  • Uses OPENCLAW_PROXY_KEY to protect public bridge APIs.

  • Includes /admin management console.

  • No runtime npm dependencies are required.

    • Environment

    Copy the example file:

    ``bash
    cp .env.example .env
    chmod 600 .env
    `

    Important values:

    `bash
    DEEPSEEK_API_KEY=sk-your-key
    DEEPSEEK_BASE_URL=https://api.deepseek.com
    DEEPSEEK_MODEL=deepseek-v4-pro
    OPENCLAW_PROXY_KEY=change-this-random-token
    ADMIN_TOKEN=change-this-admin-token
    OPENCLAW_HOME=/root/.openclaw
    `

    Do not commit real API keys.

    Start Locally

    `bash
    node src/server.js
    `

    Open:

    `text
    http://127.0.0.1:8787
    http://127.0.0.1:8787/admin
    `

    CLI test:

    `bash
    node src/cli.js --once "hello"
    `

    OpenAI-compatible test:

    `bash
    curl http://127.0.0.1:8787/v1/chat/completions \
    -H "Authorization: Bearer $OPENCLAW_PROXY_KEY" \
    -H "Content-Type: application/json" \
    -d '{"model":"deepseek-v4-pro","messages":[{"role":"user","content":"hello"}]}'
    `

    CentOS 7.9 Install

    Run from the project directory:

    `bash
    sudo bash scripts/install-centos7.sh
    `

    The script installs a CentOS 7 compatible Node.js build, copies the app to /opt/weixin-ai-deepseek, creates a systemd service, and tries to install the Weixin OpenClaw plugin.

    After editing /opt/weixin-ai-deepseek/.env, restart:

    `bash
    sudo systemctl restart weixin-ai-deepseek
    curl http://127.0.0.1:8787/health
    `

    Admin Console

    Default integrated URL:

    `text
    http://SERVER_IP:8787/admin
    `

    If running the admin console in /root/ai-cmd on port 8790:

    `text
    http://SERVER_IP:8790/admin
    `

    Enable the admin service:

    `bash
    cd /opt/weixin-ai-deepseek
    bash scripts/install-admin-service.sh
    `

    View the admin token:

    `bash
    grep '^ADMIN_TOKEN=' /opt/weixin-ai-deepseek/.env
    `

    Admin features:

  • View bridge and OpenClaw Gateway status.

  • Change DeepSeek model, temperature, max tokens, and timeout.

  • Update DeepSeek API key, proxy key, and admin token.

  • Edit OpenClaw workspace persona files.

  • Create, edit, enable, disable, and delete local skills.

  • Run limited OpenClaw skills commands: list, search, verify, install.

  • Edit /root/.openclaw/openclaw.json and detected model fields.

  • Restart weixin-ai-deepseek and openclaw-gateway.

  • Read journal logs.

    • Use HTTPS and a strong ADMIN_TOKEN before exposing /admin to the public internet.

    Weixin Plugin

    Install or update:

    `bash
    npx -y @tencent-weixin/openclaw-weixin-cli@latest install
    `

    OpenAI-compatible config:

    `text
    Provider: OpenAI Compatible
    Base URL: http://127.0.0.1:8787/v1
    API Key: OPENCLAW_PROXY_KEY from .env
    Model: deepseek-v4-pro
    `

    Useful Commands

    `bash
    sudo systemctl restart weixin-ai-deepseek
    sudo systemctl restart openclaw-gateway
    sudo journalctl -u weixin-ai-deepseek -f
    sudo journalctl -u openclaw-gateway -f
    bash scripts/install-openclaw-weixin.sh
    bash scripts/test-local.sh
    ``