Agents365-ai

Agents365-ai / pi-plugin-cc

Public

Claude Code plugin that routes /pi:review, /pi:rescue, etc. through the Pi coding agent (default: DeepSeek V4). 1:1 fork of codex-plugin-cc.

claude-code claude-code-plugin code-review coding-agent deepseek
15
0
85% credibility
Found May 26, 2026 at 15 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
JavaScript
AI Summary

This is a plugin that adds extra code review superpowers to Claude Code. When installed, it lets you call on the Pi coding agent to review your code, challenge your design decisions, or tackle complex debugging tasks. You can connect it to many different AI services (DeepSeek, OpenAI, Anthropic, Google, or local models) and run reviews either immediately or in the background. The plugin provides simple commands like /pi:review for standard reviews, /pi:adversarial-review for deeper critique, and /pi:rescue for delegating difficult problems. It's open source, well-documented, and runs entirely on your machine—no external servers involved.

How It Works

1
💡 You want smarter code reviews

You're using Claude Code to write code and wish you had a second AI assistant to double-check your work before shipping.

2
🔌 You install the plugin

You add this plugin to Claude Code with a simple command, and it connects your coding assistant to the Pi coding agent.

3
🔗 You connect your favorite AI service

You choose which AI provider you want to use—DeepSeek, OpenAI, Anthropic, Google, or even a model running on your own computer.

4
👀 You run your first code review

With a single command, Pi reviews your code changes and gives you structured feedback on what looks good and what needs attention.

5
You explore different review styles
✅
Standard review

Pi checks your code for bugs, style issues, and best practices

🔥
Adversarial review

Pi questions your approach and suggests fundamentally different solutions

6
🚀 You delegate complex tasks

For tricky bugs or big refactors, you hand the work off to Pi and let it investigate while you do something else.

🎉 Your code gets better reviews

You ship more confident code because you have a tireless AI reviewer that catches issues and explores solutions you hadn't considered.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 15 to 15 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 pi-plugin-cc?

pi-plugin-cc is a Claude Code plugin written in JavaScript that delegates code reviews and coding tasks to the Pi coding agent. It exposes slash commands like `/pi:review`, `/pi:adversarial-review`, and `/pi:rescue` directly in your Claude Code workflow. Rather than being tied to a specific LLM, the plugin routes requests through Pi, which supports DeepSeek, OpenAI, Anthropic, Google, and local options like Ollama or LM Studio via configuration. The project is a direct fork of OpenAI's codex-plugin-cc with the runtime swapped from Codex to Pi.

Why is it gaining traction?

The main hook is model freedom: you get Claude Code's CLI interface while routing requests to whichever provider you prefer, with DeepSeek V4 as the sensible default. Unlike the official Codex plugin, this requires no OAuth login—Pi authenticates via API key, which simplifies setup for teams already using multiple providers. The adversarial review feature is particularly useful for catching design-level issues rather than surface-level bugs, and the ability to run tasks in the background with status tracking fills a gap for longer investigations. Since it's model-agnostic, you can benchmark different providers against the same codebase without switching tools.

Who should use this?

Developers who want Claude Code's slash command experience but prefer using DeepSeek, Anthropic, or local models instead of being locked into OpenAI's ecosystem will find this most valuable. Teams doing structured code reviews as part of their workflow—particularly those already using Pi for other tasks—will benefit from the consistency across tools. It's also useful for Claude Code users who want to avoid Codex's OAuth requirements or who want to experiment with adversarial review as a quality gate.

Verdict

This is a niche but solid plugin for developers already invested in the Pi ecosystem or those seeking provider flexibility within Claude Code. With 15 stars and early-stage documentation, it's a credible fork with Apache 2.0 licensing and a clear upgrade path from the official codex-plugin-cc. The credibility score of 0.8500000238418579% reflects a well-documented project with proper attribution to its parent, though limited community activity means you should verify it meets your needs before committing. Worth trying if the provider flexibility solves a real pain point for you.

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.