agentleFS
VladButacuVladButacu/agentkit/skills/repo-bootstrap/SKILL.md

repo-bootstrap

Skill0 starsChanged 5 days agoMITView on GitHub

Use when setting up or standardizing agentic development configuration in a project repository, covering the instruction file, skills layout and local overrides. Skip repositories that already follow the layout unless asked to audit them.

  • Reads credentials

5 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

2 rules. 0 are common, 2 are distinctive, and 0 are very specific to this project.

  • 0

    Common

    Most of the 5 edited versions we read keep these.

  • 2

    Distinctive

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Read what already exists: instruction files, skills directories, hooks, CI.

    line 15
  2. Distinctive:

    Run the project's verification gate.

    line 28

“Common” is measured: kept by at least 6 in 10 of the 5 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Repo bootstrap

Bring a project repository to the house layout for agent-facing configuration. The layout is project-agnostic: it says where things live, never what the project's rules are.

Inspect first

Read what already exists: instruction files, skills directories, hooks, CI. Preserve all of it; this skill fills gaps and standardizes placement, it does not rewrite project rules. Obtain the user's confirmation for anything you would delete or move.

Layout

  1. AGENTS.md at the root is the single rulebook. Keep in it only what is specific to this project; the agent's global instructions already carry the house writing, verification and process rules. Useful sections: what this project is, its non-negotiables, how it is built and verified (its gate, its commands), a skills table, and where things live.
  2. CLAUDE.md contains a single pointer line, @AGENTS.md, so Claude Code reads the same rulebook. Other tools read AGENTS.md natively.
  3. .agents/skills/<name>/SKILL.md is the only place skills are written. Codex and OpenCode read it directly. Other tools read their own directory, such as .claude/skills for Claude Code and .kiro/skills for Kiro, and each one the project uses gets a local link to .agents/skills. The repository's setup command creates these links, and .gitignore lists them. Use a relative symlink on macOS and Linux (ln -sfn ../.agents/skills .claude/skills) and a directory junction on Windows (cmd /c mklink /J .claude\skills .agents\skills), which needs no admin rights. Never commit these links, because Git for Windows checks committed symlinks out as plain text files unless symlink support is turned on. Never copy a skill into a tool-specific directory.
  4. AGENTS.local.md holds per-machine overrides. Add it to .gitignore and mention it in AGENTS.md. If the repository keeps local secrets or env files that agent worktrees need (for example .env.local), add a .worktreeinclude naming them so new worktrees stay runnable.
  5. Skills wrap the repository's own commands, typically make targets; they teach when to run what, and the targets stay the source of truth for how.
  6. Enforcement hooks the repository already has (pre-commit gates, version syncing) stay exactly as they are. This layout prescribes no hook; a repository that wants a new one adds it to its own hooks directory as a project decision.

Verify

Run the project's verification gate. Confirm from a fresh session that an agent sees the skills table and the AGENTS.md rules.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.