Neura Market
codebase-recon logo

codebase-recon

Free

Analyze git history to understand a codebase before reading any code — surfaces hotspots, bug magnets, bus factor, momentum, and high-risk files (hotspot ∩ bug-magnet) via auto-scaled analysis. Install: `python3 ~/.codex/skills/.system/skill-installer/scripts/install-skill-from-github.py --repo yuji

FreeFree tier
Type
Open Source
Visit Website GitHub (40 stars)

About codebase-recon

A coding agent skill that analyzes git history to understand a codebase before reading any code. Reveals project health, risk areas, team structure, and development momentum. It probes the repo to determine scale (small/medium/large), analyzes 7 dimensions (code hotspots, bus factor, bug magnets, team momentum, firefighting frequency, recently added files, active contributors), cross-references hotspots with bug magnets to identify high-risk files, and reports findings with actionable recommendations.

Key Features

Auto-scaled analysis based on repo size (small/medium/large)
7 parallel analysis dimensions: code hotspots, bus factor, bug magnets, team momentum, firefighting frequency, recently added files, active contributors
Cross-references hotspots with bug magnets to identify high-risk files
Reports findings with actionable recommendations
Works with 20+ coding agents (Claude Code, Cline, Cursor, GitHub Copilot, Gemini CLI, etc.)
Open source under MIT license

Pros & Cons

Pros
  • Open source and free to use
  • Works with many popular coding agents
  • Auto-scaling analysis adapts to repo size
  • Cross-referencing provides high-risk file identification
  • Actionable recommendations from the analysis
  • Quick to run (just invoke /codebase-recon in agent)
Cons
  • Requires a git repository with sufficient commit history
  • Does not analyze code content, only git history metadata
  • Installation requires agent skill setup (not a standalone tool)
  • May not be useful for very small or new repositories with little history

Best For

Understanding a codebase before reading any codeOnboarding to a new projectIdentifying risk areas and knowledge concentration (bus factor)Assessing project health, momentum, and development patternsDetermining which files are hot and buggy for prioritization

FAQ

What dimensions does codebase-recon analyze?
It analyzes 7 dimensions: Code Hotspots, Bug Magnets, High-Risk Files, Bus Factor, Team Momentum, Firefighting Frequency, and Recently Added Files.
How do I install codebase-recon?
You can install via npx skills add yujiachen-y/codebase-recon-skill, or via Claude Code's plugin system or Codex's plugin system.
What are the requirements?
Git (any version) and a git repository with commit history.
Is it free?
Yes, it is open source under the MIT license.