QQQingyu

QQQingyu / feishu-codex-bridge

Public

Run local Codex CLI from Feishu/Lark chat, with sessions, attachments, and background service support.

14
2
85% credibility
Found May 23, 2026 at 19 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
TypeScript
AI Summary

Feishu-Codex Bridge is a local tool that lets you chat with OpenAI's Codex coding assistant directly from Feishu/Lark workplace messaging. You send messages to a bot in your chat app, and the bot runs Codex on your own computer—reading your project files, making edits, running terminal commands—and streams the results back to you in real-time as interactive chat messages. It's designed for developers who want AI coding help without leaving their workflow, supporting private chats, group chats (where it responds when mentioned), and even cloud document comments. The tool runs as a background service on your Mac and persists conversations per chat or topic thread.

How It Works

1
💬 You hear about chatting with your coding assistant from Feishu

A colleague mentions you can talk to an AI that works on your actual code files, right from your workplace chat app.

2
📦 You install the bridge on your computer

You run a simple install command and the tool sets itself up in a private folder on your machine, ready to run in the background.

3
📱 You scan a QR code to connect your Feishu account

The setup shows a QR code that you scan with your Feishu app, linking your account to the assistant without sharing passwords.

4
🤖 You send a message to your new coding assistant bot

You open a chat with the bot and ask something like 'Help me understand this project' or 'Add a login feature here'.

5
⚙️ The bot springs to life on your computer

Behind the scenes, the bot runs OpenAI's coding tool on your local machine, reading your actual project files and understanding your codebase.

6
💭 You watch the response stream back in real-time

As the assistant thinks and works, you see its thoughts and actions appear in your chat—reading files, making edits, running commands—like magic.

✨ Your code is updated and your question is answered

The work is done! Your files have been changed, your questions are answered, and everything happened right from your familiar chat window.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 19 to 14 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is feishu-codex-bridge?

feishu-codex-bridge is a personal bot that runs on your local machine and lets you interact with OpenAI's Codex CLI directly from Feishu or Lark chat. You send messages in a chat, the bridge streams them to a locally running Codex process, and the AI's response comes back as a streaming card or message. It handles sessions so each conversation keeps context, manages file and image attachments by downloading them locally before passing to Codex, and integrates with Lark CLI so the AI can actually call Feishu APIs to read docs, send messages, or manage calendars. The project is built in TypeScript and runs as a long-lived process connected via WebSocket to Feishu's event system.

Why is it gaining traction?

The hook is simple: you get a capable coding assistant in your existing chat tool without leaving your workflow. Developers who live in Feishu can ask Codex to inspect a repo, write code, or run commands by just typing messages instead of switching terminals. The bridge manages CLI installation privately in your home directory, so it works even on managed company machines without root access. Per-chat sessions, workspace management, and built-in commands like `/new`, `/resume`, and `/cd` make it practical for real work rather than demo purposes.

Who should use this?

This is for developers already using Feishu or Lark who want to run a local AI coding assistant without context-switching. Teams evaluating local AI coding tools will find the chat-based UX lower friction than terminal usage. Individual developers who work across multiple projects can maintain separate sessions per chat or topic group, which helps when context matters.

Verdict

The 0.85 credibility score reflects a solid engineering foundation, but the project is young with only 14 stars and limited third-party validation. Documentation is thorough and onboarding is guided, but test coverage is unclear and maturity is early. Worth trying if you live in Feishu and want Codex in your chat workflow; treat it as a personal tool rather than a team-critical system until it sees more production use.

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.