Skip to content

JPeetz/agent-skills / skills/development/repo-recon

repo-recon

Verified

Use this skill when you need to map an unfamiliar codebase before changing anything. Runs a systematic five-step recon pass — read the map, learn the commands, trace one flow end-to-end, note conventions then match them, and write it down — so you understand architecture, build/test commands, and code conventions before making a single edit. Prevents broken builds, misfiled changes, and convention-violating PRs caused by search-before-assuming. Activates on "where do I start", "understand this codebase", "how do I build/test this", "what are the conventions here", and first-time work in any repository.

FileKindSize
CHANGELOG.mdmarkdown1.3 KB
LICENSEother1.0 KB
SKILL.mdskill md8.4 KB
evals/evals.jsonconfig1.9 KB

CHANGELOG.md

View on GitHub
# CHANGELOG — repo-recon

## v1.0.0 — 2026-09-14

### Added
- Initial release of the Repo Recon agent skill (packaged from
  adityaarakeri/senior-agent-skills)
- Systematic five-step recon pass: Read the map, Learn the commands, Trace
  one flow end-to-end, Note conventions then match them, Write it down
- Structured SKILL.md sections: Overview, When to Use, The Recon Pass,
  Standing Rules, Definition of Done, Common Mistakes, Cross-reference
- Standing rules incl. search-before-assuming and verify-before-believing,
  with code kept strictly read-only during the pass
- Definition of Done checklist with explicit zero-source-changes gate
- Cross-reference to anti-over-engineering (record deferred cleanups, never
  start them during recon)
- Trigger-friendly description starting with "Use this skill when"
- 5 eval cases with near-miss negative
- MIT License (c) 2026 JPeetz

### Why
Most agent failures on unfamiliar repos come from skipping reconnaissance:
guessed commands that break the build, edits that violate local conventions,
and changes to the wrong module. Existing onboarding content is either
tool-specific or too heavy. This skill packages the lightweight senior-agent
recon loop into a triggerable, portable SKILL.md with a hard gate: no source
changes until the map, commands, flow trace, and conventions are written down.
finishing-a-development-branchobra/superpowersUse when implementation is complete, all tests pass, and you need to decide how to integrate the workStars: 323kUpdated Claude CodeCodexVerifiedInstall
systematic-debuggingobra/superpowersUse when encountering any bug, test failure, or unexpected behavior, before proposing fixesStars: 323kUpdated Claude CodeCodexVerifiedInstall
test-driven-developmentobra/superpowersUse when implementing any feature or bugfix, before writing implementation codeStars: 323kUpdated Claude CodeCodexVerifiedInstall
skill-creatoranthropics/skillsCreate new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy.Stars: 198kUpdated Claude CodeCodexVerifiedInstall