Claude Skill
pwrdrvr/openclaw-codex-app-server
OpenClaw 插件,通过 Codex 应用服务器协议将 OpenAI Codex 引入 Telegram 和 Discord。使用此 TypeScript 插件在聊天平台中启用 AI 代码生成。
概览
仓库信息
安装这个 Skill
git clone https://github.com/openclaw/openclaw.gitRegistry 信息
git clone https://github.com/openclaw/openclaw.gitpnpm add -D openclaw@file:/absolute/path/to/openclaw
项目简介
OpenClaw 插件,通过 Codex 应用服务器协议将 OpenAI Codex 集成到 Telegram 和 Discord 中,直接在聊天平台内实现 AI 驱动的代码生成和辅助。
OpenClaw plugin that brings Codex to Telegram and Discord via the Codex App Server protocol.
要点
- 与 Telegram 和 Discord 无缝集成
- 利用 Codex 应用服务器协议实现可靠通信
- 作为 OpenClaw 插件构建,易于部署
- 在聊天界面内实现 AI 代码生成
- 使用 TypeScript 编写,便于维护
使用场景
- 通过 Telegram 机器人获取代码片段和解决方案
- 将 Codex 驱动的编码助手集成到 Discord 服务器中
- 在聊天频道中自动提供代码审查建议
- 为社区成员提供按需编程帮助
- 使用 OpenClaw 原型设计和测试代码生成工作流
README 摘要
# OpenClaw Plugin For Codex App Server [](https://github.com/pwrdrvr/openclaw-codex-app-server/actions/workflows/ci.yml) [](https://www.npmjs.com/package/openclaw-codex-app-server) [](https://www.npmjs.com/package/openclaw-codex-app-server) [](./LICENSE) <p align="center"> <a href="https://youtu.be/GKkipfNEJJQ"> <img src="https://img.youtube.com/vi/GKkipfNEJJQ/maxresdefault.jpg" alt="Watch the OpenClaw Codex App Server demo on YouTube" width="100%" /> </a> </p> This project has no product name. It is just an OpenClaw plugin that connects OpenClaw to the Codex App Server protocol so you can interact with your existing threads from Codex Desktop and Codex TUI through Telegram and Discord conversations. `Codex` is mentioned here only to describe the protocol and toolchain this plugin connects to. This repository is independent and is not official, provided, sponsored, endorsed, or affiliated with OpenAI or Codex. If `codex` already works on the machine running OpenClaw, this plugin should work too. It uses the same local Codex CLI and shared login state. There is no separate plugin login requirement for normal use. ## Quick Start 1. Install the plugin into OpenClaw. 2. Start in the Telegram or Discord conversation where you want the bridge bound. 3. Run `/cas_resume`. 4. Pick a recent thread, click `New` to start a fresh one, or search directly. 5. Once bound, plain text in that conversation routes to the selected Codex thread. Buttons are presented for proje