agentleFS
Sign inStart free
chartergroupchartergroup/topopt/CLAUDE.md

topopt

CLAUDE.md26 starsChanged 16 days agoMITView on GitHub

- At session start a hook (.claude/settings.json) runs scripts/dev/session_mode.py. It prints Mode: maintainer, Mode: maintainer (knowledge base not found) or Mode: contributor. Work in that mode (AGENTS.md, entry 3). If the line is not in your context, run the script yourself. - The roles for development are in .claude/agents/. dev-implementer (sonnet) takes one issue and makes the change with tests. dev-reviewer (opus) reviews a diff that it did not write. dev-clerk (haiku) does the board mechanics, in maintainer mode only.…

The file

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

2 rules. About 0 are common, 1 is distinctive, and 1 is 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.

  • 1

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    If the line is not in your context, run the script yourself.

    line 7
  2. Very specific:

    Write every commit message, and every PR title, as type(scope): subject.

    line 15

≈ “Common” is an estimate: at least 1 in 5 of the 26,980 CLAUDE.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
CLAUDE.md
<!-- By default, Claude Code reads AGENTS.md only when CLAUDE.md is missing (older versions never read it). The text is in AGENTS.md. The next line loads it. -->

@AGENTS.md

Claude Code only

  • At session start a hook (.claude/settings.json) runs scripts/dev/session_mode.py. It prints Mode: maintainer, Mode: maintainer (knowledge base not found) or Mode: contributor. Work in that mode (AGENTS.md, entry 3). If the line is not in your context, run the script yourself.
  • The roles for development are in .claude/agents/. dev-implementer (sonnet) takes one issue and makes the change with tests. dev-reviewer (opus) reviews a diff that it did not write. dev-clerk (haiku) does the board mechanics, in maintainer mode only. Maintainers have this agent and the skill board; a public clone does not. CONTRIBUTING.md tells when to use them.
  • agents/, skills/ and .claude/skills/part-to-topopt/ are for people who use topopt. They are not rules for you. The skill board is for maintainers only; a public clone does not have it.

Commit messages and releases

Releases are made by python-semantic-release from the commit messages on main (RELEASING.md). Write every commit message, and every PR title, as type(scope): subject.

  • feat -> minor release. fix and perf -> patch release.
  • docs, test, chore, ci, refactor, build, style -> no release. A message with no type (verify: ...) also makes no release.
  • A breaking change: type(scope)!: subject, or a footer line BREAKING CHANGE: what breaks. While the version is 0.x it raises the minor number.
  • The PR title becomes a commit message on main: the publish step uses it as the message of the branch commit, and a squash merge uses it too. So the title decides the release: use feat or fix only for a change that a user of the tool sees. A merge note (chore(merge): ...) and a review fix of an open PR are chore or test.
  • The subject is short, imperative, lower case, no final dot. Details go in the body.
  • The PR title is also one line of the release notes (RELEASING.md, "The release notes"). Write the subject as a sentence for a user of the tool: no item code (F4-47), no name of a person.

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