Skip to main content

MCP server bridging Claude to Antigravity (agy), Codex, Kimi, and Claude via CLI

Project description

ai-cli-commander

MCP server that lets any AI agent orchestrate multiple LLM CLIs — Antigravity (agy), Codex, Kimi, and Claude — through a single unified interface.

One agent to rule them all. Ask GPT, Kimi, Gemini, and Claude in parallel, from any MCP-compatible host (Claude Code, OpenClaw, Cursor, etc.).


What is this?

ai-cli-commander 是一个 MCP Server 集合,把四个主流 AI 的 CLI 工具包装成统一的 MCP 工具,让任意 AI Agent 主脑可以跨模型调度:

MCP 工具 背后的 AI 最佳场景
ask_agy Antigravity / Gemini Flash Google 生态、长上下文
ask_codex OpenAI Codex / GPT-5.4 代码生成、ChatGPT 订阅免费额度
ask_kimi Kimi K2.5(月之暗面) 中文代码/文档、国内网络
ask_claude Claude Sonnet 4.6 复杂推理、多文件编辑、工具调用
ask_all 以上全部并发 并行获取多模型意见

典型用法:让 Claude Code 同时咨询 GPT 和 Kimi,比对答案;或让 Cursor 调用 Claude 做二次审查。


Architecture

┌─────────────────────────────────────────────────────────────┐
│  Agent Host(Claude Code / Cursor / OpenClaw / 任意 MCP 客户端)│
│  负责推理、规划、决定调用哪个外部 AI                             │
└─────────────────────────────────────────────────────────────┘
                             │  MCP Protocol
                             ▼
┌─────────────────────────────────────────────────────────────┐
│              ai-cli-commander MCP Server                    │
│  ask_agy() │ ask_codex() │ ask_kimi() │ ask_claude()        │
│                    ask_all()(并发)                         │
└─────────────────────────────────────────────────────────────┘
        │           │           │           │
        ▼           ▼           ▼           ▼
  ┌─────────┐ ┌─────────┐ ┌─────────┐ ┌─────────────┐
  │   agy   │ │  Codex  │ │  Kimi   │ │   Claude    │
  │--print  │ │app-svr  │ │ ACP CLI │ │ -p --json   │
  │ stdout  │ │JSON-RPC │ │JSON-RPC │ │ CLI stdout  │
  └─────────┘ └─────────┘ └─────────┘ └─────────────┘

四个 AI 的底层协议差异

AI 启动命令 协议
agy agy --print CLI stdout
Kimi kimi acp ACP (JSON-RPC over stdio)
Codex codex app-server app-server (JSON-RPC over stdio)
Claude claude -p --output-format json --bare CLI stdout JSON

注意:旧版 Gemini CLI(gemini --acp)自 2026年6月18日起对个人 Pro 用户停止服务,已由 Antigravity CLI (agy) 替代。


Prerequisites

1. Antigravity CLI (agy)

# 首次运行时按提示通过浏览器完成 OAuth 授权
agy

2. Codex CLI

npm install -g @openai/codex
codex login

3. Kimi CLI

# macOS / Linux
curl -fsSL https://code.kimi.com/install.sh | bash
# 或
uv tool install --python 3.13 kimi-cli

kimi --version
kimi /login

4. Claude Code

npm install -g @anthropic-ai/claude-code

5. uv(Python 包管理器)

curl -LsSf https://astral.sh/uv/install.sh | sh

Installation

git clone https://github.com/HiJiangChuan/ai-cli-commander.git
cd ai-cli-commander

# 安装统一包(一次安装,获得所有命令)
uv tool install .

安装后命令出现在 ~/.local/bin/,可直接执行,不依赖项目目录。

更新:

cd ai-cli-commander && git pull
uv tool install --force .

卸载:

uv tool uninstall ai-cli-commander

Register MCP Server

~/.claude.json 中注册(Claude Code 用户):

{
  "mcpServers": {
    "ai-cli-commander": {
      "command": "ai-cli-commander-server"
    }
  }
}

完整配置示例见 mcp-config-example.json

OpenClaw / Cursor 等 MCP 兼容客户端同理挂载。


Usage

注册完成后,Agent 主脑可以调用:

ask_agy("用一句话解释这个函数。")
ask_codex("把这段代码重构成 async/await 风格。")
ask_kimi("分析这个中文文档的核心观点。")
ask_claude("审查这个方案的安全性。")

# 并发询问所有 AI,对比答案
ask_all("这个算法的时间复杂度是多少?")

Project Structure

ai-cli-commander/
├── pyproject.toml                 # 统一包配置
├── README.md
├── mcp-config-example.json        # MCP 注册配置示例
│
└── src/ai_commander/
    ├── __init__.py
    ├── server.py                  # 统一 MCP Server 入口
    ├── core.py                    # 公共代码:ACPClient、LineBuffer、日志、CLI 检查
    ├── agy.py                     # Antigravity handler
    ├── codex.py                   # Codex handler
    ├── kimi.py                    # Kimi handler
    └── claude.py                  # Claude handler

License

MIT

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

ai_cli_commander-1.0.0.tar.gz (14.1 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

ai_cli_commander-1.0.0-py3-none-any.whl (18.7 kB view details)

Uploaded Python 3

File details

Details for the file ai_cli_commander-1.0.0.tar.gz.

File metadata

  • Download URL: ai_cli_commander-1.0.0.tar.gz
  • Upload date:
  • Size: 14.1 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: uv/0.11.2 {"installer":{"name":"uv","version":"0.11.2","subcommand":["publish"]},"python":null,"implementation":{"name":null,"version":null},"distro":{"name":"macOS","version":null,"id":null,"libc":null},"system":{"name":null,"release":null},"cpu":null,"openssl_version":null,"setuptools_version":null,"rustc_version":null,"ci":null}

File hashes

Hashes for ai_cli_commander-1.0.0.tar.gz
Algorithm Hash digest
SHA256 f9bcb1b6ced66a2eccccbee514a0d4c0e4b5945aa22bb310aec3d63c8f6328d3
MD5 460cfbab85bd8cb9478e52be5965f6e2
BLAKE2b-256 ca992554a23de2e189b4b9d09b8fcc3877f93862fedb8688a09989c5c3d91b16

See more details on using hashes here.

File details

Details for the file ai_cli_commander-1.0.0-py3-none-any.whl.

File metadata

  • Download URL: ai_cli_commander-1.0.0-py3-none-any.whl
  • Upload date:
  • Size: 18.7 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: uv/0.11.2 {"installer":{"name":"uv","version":"0.11.2","subcommand":["publish"]},"python":null,"implementation":{"name":null,"version":null},"distro":{"name":"macOS","version":null,"id":null,"libc":null},"system":{"name":null,"release":null},"cpu":null,"openssl_version":null,"setuptools_version":null,"rustc_version":null,"ci":null}

File hashes

Hashes for ai_cli_commander-1.0.0-py3-none-any.whl
Algorithm Hash digest
SHA256 cd8787dbb25ece98d451f16ce11aa9037a445eac1de48227a98340fa4ce91556
MD5 023a1dce34f593025130ebbae1fd1db5
BLAKE2b-256 3aacfd63be9f584b37ae69e5b60b52671172cf4d50ea31711cb2238f9c6ccee3

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page