AgentSkillsCN

00 Bootstrap Repo

00 引导式仓库

SKILL.md
--- frontmatter
tags: [skill, codex-exec]
skill_id: "00_bootstrap_repo"

Skill 00 — Bootstrap Repo Overview

Purpose: Generate a lightweight repo overview (tree + stats + entrypoints) so other skills can work without loading huge context.

Protocols:

  • [[docs/PROTOCOLS/VAULT_PROBE]]
  • [[docs/PROTOCOLS/EVIDENCE]]
  • [[docs/PROTOCOLS/RUN_LOGGING]]
  • [[docs/PROTOCOLS/PATCH_GATE]]
  • [[docs/PROTOCOLS/CONTEXT_BUDGET]]

Tools:

  • [[tools/repl/README]] — patch-gated edits (generate + check + apply)

Inputs (from Orchestrator)

  • Task context (optional): what kind of project this is (web app, API, scripts, etc.)
  • Any known entrypoints (optional)

Outputs (must write)

  • artifacts/repo_tree.txt
  • artifacts/repo_stats.json
  • docs/REPO_OVERVIEW.md
  • [[docs/skills/00_bootstrap_repo/SUMMARY]]

Run log (mandatory)

Create: artifacts/_runs/<timestamp>_00_bootstrap_repo/

Include:

  • COMMANDS.md (all commands executed)
  • LOG.txt (combined stdout/stderr)
  • NOTES.md (short notes)

Steps

  1. Run a shallow tree (depth 4–6) to understand layout:
    • node tools/probe/probe.mjs tree --root . --depth 5
  2. Generate repo stats:
    • node tools/probe/probe.mjs stats --root . --json
  3. Identify likely entrypoints using probe (examples):
    • package.json, README.md, apps/, src/, main.*, index.*
  4. Update docs/REPO_OVERVIEW.md:
    • High-level structure
    • How to run/build/test (if discoverable)
    • Key modules/entrypoints
  5. Write SUMMARY.md with evidence pointers and the commands used.

Definition of Done

  • All required output files exist
  • docs/REPO_OVERVIEW.md includes: structure + how-to-run + key entrypoints (or explicit “not found” with evidence)
  • SUMMARY.md includes evidence pointers (path:Lx-Ly) and probe commands