AutoReview

A 3-stage AI review pipeline that analyses any codebase using Claude Opus 4.6 and Codex 5.3 and produces a single, arbitrated "supreme report".

How it works

Your query
  │
  ├── Stage 1 (parallel)   Claude + Codex independently analyse the codebase
  ├── Stage 2 (parallel)   Each model cross-reviews the other's findings
  └── Stage 3              Codex arbitrates both reports → supreme report

The tool is read-only — it never modifies the codebase being analysed.

Requirements

• Python 3.10+
• claude CLI installed and authenticated
• codex CLI installed and authenticated
• pip install rich

Usage

pip install rich
cd /path/to/any/codebase
python /path/to/AutoReview/autoreview.py

Enter a query at the prompt, e.g.:

Enter your query: summarize the main purpose of this codebase

The report is saved as review_YYYYMMDD_HHMMSS.md in the current working directory.