codebase-recon-skill

April 26, 2026 ยท View on GitHub

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.

Inspired by "The Git Commands I Run Before Reading Any Code" by Ally Piechowski.

Installation

Via skills.sh (works with 20+ coding agents)

npx skills add yujiachen-y/codebase-recon-skill

Works with Claude Code, Cline, Cursor, GitHub Copilot, Gemini CLI, and any agent supporting the Agent Skills Specification.

Via Claude Code plugin system

/plugin marketplace add yujiachen-y/codebase-recon-skill

Then install the plugin from the marketplace browser via /plugin.

Via Codex plugin system

codex plugin marketplace add yujiachen-y/codebase-recon-skill

Then run /plugins in Codex, choose the Codebase Recon marketplace, and install codebase-recon.

Usage

In your coding agent, invoke:

/codebase-recon

The skill will:

  1. Probe the repo to determine its scale (small / medium / large)
  2. Analyze 7 dimensions in parallel: code hotspots, bus factor, bug magnets, team momentum, firefighting frequency, recently added files, and active contributors
  3. Cross-reference hotspots with bug magnets to identify high-risk files
  4. Report findings with actionable recommendations

What You'll Learn

DimensionQuestion Answered
Code HotspotsWhich files change the most?
Bug MagnetsWhich files attract the most bug fixes?
High-Risk FilesWhich files are both hot AND buggy?
Bus FactorWho knows what? Is knowledge concentrated?
Team MomentumIs development accelerating, stable, or declining?
FirefightingHow often are there emergency fixes and reverts?
Recently AddedWhere is active development happening?

Requirements

  • git (any version)
  • A git repository with commit history

Attribution

This skill is inspired by "The Git Commands I Run Before Reading Any Code" by Ally Piechowski. The original article describes 5 git commands for codebase reconnaissance. This skill extends the concept with auto-scaling, cross-referencing, and actionable recommendations.

License

MIT