agentleFS
Sign inStart free
rajfnurajfnu/agent-engineering-standard/AGENTS.md

agent-engineering-standard

AGENTS.md0 starsChanged 51 days agoApache-2.0View on GitHub

Instructions for agents contributing to the Agent Engineering Standard itself. For the standard's guidance on writing software generally, see skills/. A collection of portable Agent Skills. The product is instructions, not code. There is more Markdown here than source, and that is correct — do not "add an implementation". 1. skills/ is the single source of truth. Adapters (.claude-plugin/, .codex-plugin/, commands/) point at it and must never contain engineering instructions. If you find yourself copying a rule into a command,…

  • 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 31,774 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 repository4 other files it gives its agents.

The file

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

4 rules. About 0 are common, 4 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈4

    Distinctive · estimate

    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:

    No rule without a failure: Every rule must name the concrete failure it prevents.

    line 18
  2. Distinctive:

    If you cannot state it, do not add it.

    line 19
  3. Distinctive:

    Stay agnostic: No language, framework, build tool, or single agent assumed.

    line 22
  4. Distinctive:

    If skill sizes changed, refresh the published measurement tables:

    line 32

≈ “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

Working on this repository

Instructions for agents contributing to the Agent Engineering Standard itself. For the standard's guidance on writing software generally, see skills/.

What this repository is

A collection of portable Agent Skills. The product is instructions, not code. There is more Markdown here than source, and that is correct — do not "add an implementation".

Rules

  1. skills/ is the single source of truth. Adapters (.claude-plugin/, .codex-plugin/, commands/) point at it and must never contain engineering instructions. If you find yourself copying a rule into a command, stop.
  2. Respect the budgets. SKILL.md stays under 200 lines; detail goes to references/. CI enforces this.
  3. No rule without a failure. Every rule must name the concrete failure it prevents. If you cannot state it, do not add it.
  4. Do not duplicate integrated tools. OpenCodeReview owns line-level review; Superpowers owns workflow methodology. See docs/decisions/.
  5. Stay agnostic. No language, framework, build tool, or single agent assumed.
  6. Verify external commands against upstream before documenting them. They change.

Before finishing

python3 tools/validate_skills.py skills
python3 tools/validate_repo.py

If skill sizes changed, refresh the published measurement tables:

python3 tools/measure_context.py skills --write

That rewrites the tables between the <!-- MEASUREMENTS:START --> / <!-- MEASUREMENTS:END --> markers in README.md, docs/architecture.md, and evals/RESULTS.md. validate_repo.py fails if they are stale, so this is not optional.

Scratch space

Working notes and OpenCodeReview output go in .aes/ (gitignored). Do not commit agent-generated plan or summary Markdown into the repository — that is precisely the sediment repository-reconciliation.md warns about.

Decisions

Architectural decisions live in docs/decisions/. If you are about to reverse one (build a CLI, add hooks, vendor third-party code), write an ADR superseding the existing one rather than quietly changing course.

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.

Check your setup, free