agentleFS
TheAstroByteTheAstroByte/astrobyte-agent-skills/AGENTS.md

astrobyte-agent-skills

AGENTS.md1 starsChanged 13 days agoMITView on GitHub

This repository distributes portable instructions to other agents. Treat routing metadata, completion criteria, installer safety, and cross-harness behavior as product code. If the user requests installation or adoption of this stack, or invokes $adopt-skill-stack or /adopt-skill-stack, read skills/adopt-skill-stack/SKILL.md. Default to quick installation of the committed Astrobyte profile; run the onboarding interview only when explicitly requested. Honor an explicitly selected subset. Do not treat a clone as consent to install, publish, migrate, or overwrite anything. Before changing this repository, read CONTRIBUTING.md.…

The file

We split this AGENTS.md into the rules it gives an agent

3 rules. About 0 are common, 2 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Add or update tests for installer or validator behavior.

    line 28
  2. Distinctive:

    Commit the verified checkpoint with a conventional message.

    line 32
  3. Very specific:

    Every skill directory has a matching lowercase kebab-case name and a useful trigger description.

    line 41

≈ “Common” is an estimate: at least 1 in 5 of the 40,246 AGENTS.md files we read have a rule of the same kind, not the same words. “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
AGENTS.md

Agent instructions

This repository distributes portable instructions to other agents. Treat routing metadata, completion criteria, installer safety, and cross-harness behavior as product code.

If the user requests installation or adoption of this stack, or invokes $adopt-skill-stack or /adopt-skill-stack, read skills/adopt-skill-stack/SKILL.md. Default to quick installation of the committed Astrobyte profile; run the onboarding interview only when explicitly requested. Honor an explicitly selected subset. Do not treat a clone as consent to install, publish, migrate, or overwrite anything.

Repository contribution workflow

Before changing this repository, read CONTRIBUTING.md. Work on a feature branch and submit a pull request against main. Only the repository owner, TheAstroByte, merges pull requests. Agents leave merging to the owner, including when using the owner's credentials. Do not push directly to main, force-push it, delete it, or enable auto-merge.

These rules govern this repository, not the reusable policies installed into other projects.

Read by task

  • Skill, AGENTS.md, or README wording: read skills/writing-for-agents/SKILL.md and its SKILL-MECHANICS.md before editing.
  • Router, profile, or skill composition: read docs/architecture.md, catalog.json, and skills/astrobyte-mode/SKILL.md.
  • Selection or replacement decisions: read docs/skill-stack-assessment.md and docs/upstreams.md.
  • Installer or harness behavior: read scripts/install.py and the relevant file under adapters/. For routing, also read docs/routing.md and scripts/routing.py.
  • Project startup or stack defaults: read skills/astrobyte-project-startup/SKILL.md, skills/astrobyte-project-startup/references/stack-policy.md, and skills/astrobyte-project-startup/references/project-defaults.json.
  • Upstream refresh: read docs/upstreams.md and preserve the matching license file.
  • Adoption or profile changes: read skills/adopt-skill-stack/SKILL.md, its relevant references, docs/adoption.md, and docs/profile-reference.md.
  • Agent-created Git commit or push: read skills/astrobyte-git-identity/SKILL.md, use its helper, and audit the unpublished range.

Change contract

  1. Inspect the current source and the affected completion criteria.
  2. Make one coherent change at a time.
  3. Add or update tests for installer or validator behavior.
  4. Run python3 scripts/validate.py.
  5. Run python3 -m unittest discover -s tests -v.
  6. Inspect the diff for secrets, machine paths, stale source pins, and accidental upstream edits.
  7. Commit the verified checkpoint with a conventional message.

Invariants

  • The default installation previews its plan and asks for confirmation. Onboarding is optional; approved unattended installation uses --yes.
  • Machine answers stay outside this repository.
  • Existing unrelated skills remain untouched.
  • Conflicting managed skills are backed up before replacement.
  • The installer changes instruction files only inside its ownership markers, and stops on anything it cannot attribute. docs/routing.md is the routing contract.
  • Every skill directory has a matching lowercase kebab-case name and a useful trigger description.
  • Owned skills keep harness-specific details behind pointers in adapters/ or skill references.
  • Vendored files retain attribution and a pinned upstream commit.
  • Documentation states durable behavior and points to executable sources for discoverable details.
  • unslop remains in every catalog profile.
  • Raw onboarding answers, secrets, hostnames, usernames, serial identifiers, local endpoints, and absolute machine paths never enter a committed operator profile.

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.