agentleFS
jingyibuildsjingyibuilds/rubber-duck/AGENTS.md

rubber-duck

AGENTS.md0 starsChanged 18 days agoMITView on GitHub

This file explains how to change Rubber Duck without breaking its package or its prompt. Rubber Duck is an agent skill written in Markdown. SKILL.md is the prompt that agents read. There is no build step and no code. Keep the skill portable. Do not write instructions that only work in one agent tool. Name tools (Claude Code, Codex, Cursor) only as examples.

  • 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 repository2 other files it gives its agents.

The file

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

5 rules. About 0 are common, 1 is distinctive, and 4 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.

  • 4

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Keep the skill portable. Do not write instructions that only work in one agent tool.

    line 9
  2. Very specific:

    Keep SKILL.md and README.md in sync: If a rule, tag, or map section changes, update the README tables and the example.

    line 23
  3. Very specific:

    Version: keep the same version in SKILL.md under metadata.version, the first CHANGELOG.md heading, .claude-plugin/plugin.json, and …

    line 24
  4. Very specific:

    Length: every word of SKILL.md is read on every use.

    line 26
  5. Very specific:

    Instructions, not decoration: every line in SKILL.md should tell the agent to do something specific at a specific moment.

    line 27

≈ “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

Guide for agents

This file explains how to change Rubber Duck without breaking its package or its prompt.

What this repo contains

Rubber Duck is an agent skill written in Markdown. SKILL.md is the prompt that agents read. There is no build step and no code.

Keep the skill portable. Do not write instructions that only work in one agent tool. Name tools (Claude Code, Codex, Cursor) only as examples.

Key files

  • SKILL.md: the source of truth and the only skill file.
  • README.md: what the skill does, installation, usage. Its Before/After example must follow the map format in SKILL.md.
  • docs/DECISIONS.md: why the skill is shaped the way it is, including the options that were rejected. Add an entry for every design change.
  • CHANGELOG.md: release notes, newest first.
  • .claude-plugin/plugin.json and .claude-plugin/marketplace.json: the Claude plugin and marketplace entry. The skill loader points at the root SKILL.md.
  • .cursor-plugin/plugin.json: the Cursor plugin.
  • agents/openai.yaml: display name, short description, and default prompt for OpenAI-compatible agents.

Rules for changes

  • Keep SKILL.md and README.md in sync. If a rule, tag, or map section changes, update the README tables and the example.
  • Version: keep the same version in SKILL.md under metadata.version, the first CHANGELOG.md heading, .claude-plugin/plugin.json, and .cursor-plugin/plugin.json.
  • Description: the plugin manifests use the first sentence of the SKILL.md description.
  • Length: every word of SKILL.md is read on every use. Keep the body (everything below the frontmatter) under 1,600 words. A change that adds words has to earn them, and a new rule should replace an old one where it can.
  • Instructions, not decoration: every line in SKILL.md should tell the agent to do something specific at a specific moment. "Be thoughtful" is not an instruction. "Ask at most one question per reply" is.
  • Evidence: a new rule needs a reason in docs/DECISIONS.md: an observed failure, a user's reaction, or a source.

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.