Back to browse
GitHub Repository

Automatically analyzes your codebase to learn conventions, then guides AI-generated code to follow team idioms, architectural patterns, and best practices — keeping auto-generated code in sync with evolving standards.

9 starsPython

Chameleon – shows Claude a real file from your repo before it edits

by crisnahine·Jul 12, 2026·1 point·0 comments

AI Analysis

●●●BangerBig BrainSolve My ProblemNiche Gem

Injects real repo witnesses and counterexamples before edits to stop AI style drift.

Strengths
  • PreToolUse hooks inject actual conforming files rather than generic style guidelines.
  • Counterexamples flag specific anti-patterns found in your own codebase automatically.
  • Maps inbound callers to ensure signature changes update call sites in the same turn.
Weaknesses
  • Currently limited to Claude Code, excluding Copilot, Cursor, and other agent runtimes.
  • Relies on successful archetype detection which may struggle with highly polyglot repos.
Target Audience

Engineering teams using Claude Code for automated refactoring or feature generation

Similar To

Cursor Rules · Continue.dev · Sourcegraph Cody

Similar Projects