agentleFS
Sign inStart free
LeostrukaLeostruka/devin-bundle/skills/project-bootstrap/templates/agents.md

devin-bundle / templates

AGENTS.md0 starsChanged 3 months agoMITView on GitHub

This file is included in the agent's context for this project. Keep it concise; only rules that change the agent's behavior belong here. Generic rules live in the global bundle.

The file

Show its rules and text

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

13 rules. About 3 are common, 10 are distinctive, and 0 are very specific to this project.

  • ≈3

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈10

    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
  • could be automatic: a setting could enforce it
  1. Distinctive:

    If uncertain, ask — don't guess.

    line 7
  2. Distinctive:

    Use the smallest solution that solves the problem.

    line 9
  3. Distinctive:

    Run linters after finishing a task.

    line 16
  4. Distinctive:

    Don't delete, disable, or skip tests without explicit approval.

    line 21
  5. Distinctive:

    Run the test suite before declaring a task done.

    line 22
  6. Common:

    Never output or log secrets, tokens, passwords, or API keys.

    line 26
  7. Common:

    Never commit secrets to the repository.

    line 27
  8. Distinctive:

    Treat user input as untrusted.

    line 28
  9. Distinctive:

    Default endpoints, storage URLs, and services to private.

    line 29
  10. Common:

    Use secure defaults: confirmations for destructive actions, hidden credentials, least privilege.

    line 30
  11. Distinctive:

    Keep PRs around 300 lines of change.

    line 34
  12. Distinctive:

    If a task will exceed 500 lines, break it into smaller tasks with clear boundaries.

    line 35
  13. Distinctive:

    Every ticket/spec must include the intent and the "why".

    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

Project agent rules

This file is included in the agent's context for this project. Keep it concise; only rules that change the agent's behavior belong here. Generic rules live in the global bundle.

1. Intent and boundaries before code

  • If uncertain, ask — don't guess.
  • Declare the intent, user-visible impact, and boundaries (inputs/outputs, scope, non-goals) before writing code.
  • Use the smallest solution that solves the problem. Reject overengineering.

2. Code style

  • <!-- Stack-specific rules go here. Examples: - TypeScript: never use `any`; prefer strict types. - Python: follow PEP 8; use type hints where it clarifies intent. - Run linters after finishing a task. -->

3. Testing

  • Tests verify intent, not just behavior.
  • Don't delete, disable, or skip tests without explicit approval.
  • Run the test suite before declaring a task done.

4. Security

  • Never output or log secrets, tokens, passwords, or API keys.
  • Never commit secrets to the repository.
  • Treat user input as untrusted; validate and sanitize before use.
  • Default endpoints, storage URLs, and services to private. Public endpoints require documented justification.
  • Use secure defaults: confirmations for destructive actions, hidden credentials, least privilege.

5. Task size

  • Keep PRs around 300 lines of change.
  • If a task will exceed 500 lines, break it into smaller tasks with clear boundaries.
  • Every ticket/spec must include the intent and the "why".

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.

Reviews can't be read right now.

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