agentleFS
fimokleifimoklei/maestro/AGENTS.md

maestro

AGENTS.md1 starsChanged 47 days agoMITView on GitHub

Maestro is a local, visual cockpit to see and steer an AI agent setup — skills, hooks and MCP servers — across repos and tools, built on top of APM. APM is the engine, Maestro the cockpit above it (ADR-0001). Code: before writing, changing or reviewing any code, read CODING_STANDARDS.md — its gates and rule files bind every code change. - These rules outrank any always-on mode. Where an ambient ruleset conflicts with a rule below, follow this file and…

  • Where files are headingNew AGENTS.md files have more rules about keeping secrets out of the code, and fewer about code style.

    Of the AGENTS.md files we read, the 32,671 first seen in the last six months, against the older ones: the kind of rule that gained most share, and the kind that lost most. About files like this in general, not this one, and an estimate, since a kind is a pattern in the words.

  • In this repository5 other files it gives its agents.

The file

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

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

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 2

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Do the job properly, at the scope asked: No stubs, no placeholders, no half-finished paths.

    line 15
  2. Very specific:

    Docs: condense rather than expand: Before calling doc work done, check the touched docs against the conflict rule in …

    line 23
  3. Very specific:

    When the operator corrects a mistake, add its rule here with what enforces it.

    line 28

≈ “Common” is an estimate: at least 1 in 5 of the 41,581 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

AGENTS.md — Maestro

Maestro is a local, visual cockpit to see and steer an AI agent setup — skills, hooks and MCP servers — across repos and tools, built on top of APM. APM is the engine, Maestro the cockpit above it (ADR-0001).

Code: before writing, changing or reviewing any code, read CODING_STANDARDS.md — its gates and rule files bind every code change.

Hard Rules

  • These rules outrank any always-on mode. Where an ambient ruleset conflicts with a rule below, follow this file and say so.
  • Do the job properly, at the scope asked. No stubs, no placeholders, no half-finished paths. "The best version" means the best version of this job — not a bigger one.
  • Drive APM; never reimplement it. Install, sync, pinning, lockfiles and multi-tool targeting are APM's.
  • Build a new capability only through a job issue (label job, docs/operating-model.md); improving what already exists needs only a tracker issue.
  • Docs: condense rather than expand. Before calling doc work done, check the touched docs against the conflict rule in docs/operating-model.md.

Rules and their checks

When the operator corrects a mistake, add its rule here with what enforces it; a code rule goes to CODING_STANDARDS.md → Checks. A rule enforced by nothing that repeats gets a check in the same change.

RuleEnforced by
Merge only after CI is greenworkflow-ship → wait-for-checks.sh; ruleset on main
Product copy reads natural and states traced factsjudgement: .claude/rules/copy.md, copy-reviewer

Where to look

  • Designing a screen, dialog or use case → read .claude/rules/design.md whole. Drafting user-facing copy, in code, a spec, a ticket or a grill option → .claude/rules/copy.md; complete its copy review.
  • Where code lives → docs/agents/codebase-map.md. Before measuring anything → docs/research/; head -n1 docs/research/*.md lists each note's question. Domain terms → GLOSSARY.md; decisions → docs/adr/ (docs/agents/domain.md). Read LEARNINGS.md at session start.
  • Briefing a sub-agent that works in its own worktree → paste the block in docs/agents/worktree-brief.md.
  • Committing → workflow-commit. Shipping → workflow-ship. Publishing an issue (job, spec or ticket) → docs/agents/issue-tracker.md.
  • Browser: the cockpit → agent-browser (the verify-ui skill); a live third-party site → the Chrome tools (the operator's logged-in 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.