agentleFS
Sign inStart free
jabbott-iiijabbott-iii/Tidepool/CLAUDE.md

Tidepool

CLAUDE.md1 starsChanged 10 days agoApache-2.0View on GitHub

The project rules are in AGENTS.md (imported above), shared with any other coding agent. This section covers only what is specific to Claude Code. Project skills (in .claude/skills): Subagents (in .claude/agents): Each owner also keeps the README.md of the crates it changes true (ADR-0023). Path-scoped rules in .claude/rules load automatically when you open matching files: Permissions and hooks: - .claude/settings.json pre-approves routine cargo commands, including cargo xtask and cargo tp, and read-only git commands. - It denies Git operations that…

  • Reads credentials
  • Installs packages

The file

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

2 rules. About 0 are common, 2 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    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:

    Use the task list for any roadmap task with more than two steps.

    line 67
  2. Distinctive:

    When an acceptance criterion needs an OS, tool or device this session can't run (Claude Code sessions are usually Linux without the Android …

    line 71

≈ “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

@AGENTS.md

Claude Code specifics

The project rules are in AGENTS.md (imported above), shared with any other coding agent. This section covers only what is specific to Claude Code.

Project skills (in .claude/skills):

SkillUse it to…
/next-task [task-id]Pick the next unblocked roadmap task, or the one named, ask its Decide first questions, and carry it through the definition of done
/verifyRun the full verification suite and report exact results
/record-decision <title>Append an ADR to docs/decisions.md, or accept or supersede one
/evaluate-dependency <crate> <need>Run the build-vs-buy evaluation before anyone proposes a third-party crate
/add-ffi-binding <os> <api>Add OS bindings to tidepool-sys and wrap them safely in tidepool-platform
/add-managed-tool <tool>Add or update a managed tool: descriptor, detection, install plan, tests, panel row

Subagents (in .claude/agents):

AgentOwns
architectDesign, ADRs, crate boundaries, protocol and format specs (writes only in docs/)
editor-corecore, text, codec, regex, workspace, editor (including vim mode), vcs
platform-uisys, platform, gfx, font, ui, gui, tui, term, the jabbott.dev theme
language-supportsyntax, markdown, rpc, lang, language packs
mobile-toolchaintoolchain (all managed tools), mobile
ai-agentsagent, TAP, predictive text, agents/
test-engineertestkit, harnesses, xtask, bench, CI workflows
security-reviewerRead-only review; mandatory for sys/platform, toolchain, mobile, vcs, agent, parser and workflow changes

Each owner also keeps the README.md of the crates it changes true (ADR-0023).

Path-scoped rules in .claude/rules load automatically when you open matching files:

Rule fileApplies to
dependencies.mdCargo manifests and lockfiles, .cargo/, xtask, rust-toolchain.toml
unsafe-and-ffi.mdsys, platform
parsers.mdcodec, font, markdown, syntax, rpc, term, vcs, fuzz corpora
toolchains-and-secrets.mdtoolchain, mobile, vcs, toolchains/
ai-context.mdagent, agents/
ui-style.mdui, gui, tui, editor, themes/, assets/
docs-and-roadmap.mddocs/, AGENTS.md, CLAUDE.md, README.md, CONTRIBUTING.md, SECURITY.md, crate READMEs
ci-workflows.md.github/

Permissions and hooks:

  • .claude/settings.json pre-approves routine cargo commands, including cargo xtask and cargo tp, and read-only git commands.
  • It denies Git operations that publish or finalize changes, cargo publish and cargo login, and reading signing material and .env files.
  • It asks before cargo add, cargo remove, cargo update and cargo install, and before git checkout, restore, switch, stash and reset.
  • A PostToolUse hook (.claude/hooks/deps-guard.sh) runs xtask deps and then xtask layers after an edit to a manifest, a lockfile or a cargo configuration file. It falls back to scripts/check-deps.sh when xtask cannot be built (docs/04-quality.md → "Command line and rules").
  • These are guardrails, not a security boundary. The rules in AGENTS.md still apply when a command slips past a pattern.

Working style:

  • Use the task list for any roadmap task with more than two steps. Make the last item "verify", and run the task's verification commands for it.
  • Most tasks touch one or two crates. If a change spreads across more than three crates, stop and say so: the task is probably mis-scoped.
  • When an acceptance criterion needs an OS, tool or device this session can't run (Claude Code sessions are usually Linux without the Android SDK or Xcode), implement it and say plainly which checks still need it.
  • rust-toolchain.toml pins 1.99.0. If rustup cannot install it in this session, say so, and say which toolchain any command you report ran on.

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