# Repo Intake

> Onboard into an unfamiliar repository fast before making any changes

- Skill: `lucassantana-dev/repo-intake` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds add lucassantana-dev/repo-intake`
- Raw SKILL.md: https://api.skillmd.com/api/skills/lucassantana-dev/repo-intake/raw
- Safety review: pending (external: skill-scanner PASS, skillspector PASS)
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: LucasSantana-Dev (https://skillmd.com/u/lucassantana-dev)
- Updated: 2026-08-19
- Page: https://skillmd.com/skills/lucassantana-dev/repo-intake

---


# Repo Intake

Read local docs, detect the stack, summarize the structure, and flag blockers — before touching any code.

## Steps

1. **Read local instructions** — check README.md, AGENTS.md, CLAUDE.md, CONTRIBUTING.md
2. **Detect the stack** — scan package manifests, lockfiles, Dockerfiles, CI configs
3. **Map the structure** — identify entrypoints, source layout, test directories, scripts
4. **List primary commands** — run, test, build, lint, deploy (only what actually exists)
5. **Flag blockers** — missing env files, auth, broken deps, unclear setup steps

## Output

```text
Repo:      <name>
Stack:     <language, framework, package manager>
Commands:  <run | test | build | lint>
Blockers:  <missing auth | broken deps | none>
Next:      <top 1-3 actions>
```

## Rules

- Read docs before scanning code
- Never guess commands that are not present in the repo
- Stop if the repo root is unclear or ambiguous
- Flag missing setup steps instead of inventing them