agentleFS
Sign inStart free
DocSpringDocSpring/docspring-agent-playbook/AGENTS.md

docspring-agent-playbook

AGENTS.md1 starsChanged 2 days agoMITView on GitHub

Engineering principles, AGENTS.md templates and workflow skills that other teams copy and adapt. Everything here is public and meant to work for any stack, tracker and CI provider. - Plain, direct English, US spelling. Address the agent as "you" and the person as "the user". Imperative, present tense, standing instructions rather than history. - Each rule states the rule first and the reason in the same bullet. Explain the failure mode (what goes wrong and why); don't describe incidents from…

  • 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 31,439 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 repository7 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:

    No em dashes or en dashes.

    line 16
  2. Very specific:

    Keep SKILL.md short with the most important rules first.

    line 28
  3. Very specific:

    Keep one copy of the rules, in this file.

    line 36

≈ “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

DocSpring Agent Playbook

Engineering principles, AGENTS.md templates and workflow skills that other teams copy and adapt. Everything here is public and meant to work for any stack, tracker and CI provider.

Writing rules

  • Plain, direct English, US spelling. Address the agent as "you" and the person as "the user". Imperative, present tense, standing instructions rather than history.
  • Each rule states the rule first and the reason in the same bullet. Explain the failure mode (what goes wrong and why); don't describe incidents from any company's history, and no company names, ticket numbers, dates, hostnames or figures from a business.
  • Rules are opinionated. Where a rule is a policy choice that won't fit every team, add one line starting "When this doesn't fit:" with the condition and what to do instead. Rules for one kind of product go under a heading that starts with "If" ("If you generate client libraries").
  • No em dashes or en dashes. Avoid the usual AI tells: "not X but Y" reveals, forced lists of three, one-line closers that restate the paragraph, marketing adjectives.
  • Name a tool, tracker or CI provider only as a labeled example ("for example, GitHub Actions"). Skills read those details from the adopting repository's ## Agent workflow settings section (CONFIGURE.md).
  • Rules too specific to one product don't belong here. Neither does anything that describes how to get past a particular system's defences.

Skills

  • Each skill is skills/<name>/SKILL.md with name (matching the folder), description (under 1,024 characters: what it does, "Use when ..." in the user's words, "Not for ..." when there's a likely mix-up) and metadata.version.
  • Keep SKILL.md short with the most important rules first. Put long checklists in references/ and say exactly when to read each one.
  • Subagent definitions live in the skill's agents/ folder.
  • Bump the skill's metadata.version and the plugin version in .claude-plugin/ and .codex-plugin/ in the PR that changes a skill, and add a line to CHANGELOG.md.

Files

  • Keep one copy of the rules, in this file. If a tool needs a CLAUDE.md, make it a symlink to AGENTS.md or a one-line @AGENTS.md import, never a second copy (guides/writing-agents-md.md).
  • templates/ and principles/ must stay consistent: a rule changed in one changes in the other in the same PR. The reason for a rule lives once, in principles/; templates state the rule and link there.

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