agentleFS
rebornixrebornix/Agmente/CLAUDE.md

Agmente

CLAUDE.md545 starsChanged 9 months agoMITView on GitHub

This file is the entrypoint for cloud/Claude coding agents. Use it with Agents.md.

The file

Show its rules and text

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

14 rules. About 0 are common, 8 are distinctive, and 6 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈8

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 6

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Use root .agmente.paths for machine-local external repo paths.

    line 16
  2. Distinctive:

    Read this file when running upstream-spec/source checks.

    line 18
  3. Distinctive:

    Keep ACP and Codex behaviors separate unless a shared abstraction is intentional and tested.

    line 23
  4. Distinctive:

    Prefer component-local changes.

    line 26
  5. Distinctive:

    Do not add personal local filesystem paths to tracked files.

    line 27
  6. Distinctive:

    If behavior differs by protocol, document both expected paths.

    line 28
  7. Distinctive:

    Update Agents.md and this file in the same PR when:

    line 32
  8. Very specific:

    Update nested <component>/AGENTS.md in the same PR when behavior in that component changes.

    line 36
  9. Very specific:

    If no docs are needed, include Docs: N/A in PR notes.

    line 37
  10. Distinctive:

    Run focused tests for touched components first.

    line 40
  11. Very specific:

    Treat run, execute, validate, and reproduce E2E requests as execute-only and agentic: run the real backend and tools, capture evidence, …

    line 47
  12. Distinctive:

    Do not modify repo files during an execute-only E2E run unless the user explicitly asks to fix, stabilize, add hooks, or make the scenario …

    line 48
  13. Very specific:

    Do not assume the in-session mcp_xcodebuildmcp_* subset includes interaction tools (tap, swipe, gesture, type-text, long-press).

    line 53
  14. Very specific:

    If any required UI automation tool is missing from MCP exposure, switch to xcodebuildmcp CLI for simulator interaction.

    line 54

≈ “Common” is an estimate: at least 1 in 5 of the 27,641 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

Agmente Cloud Agent Guide

This file is the entrypoint for cloud/Claude coding agents. Use it with Agents.md.

Read Order

  1. Agents.md (root policy and protocol overview)
  2. Component docs closest to changed code:
    • Agmente/AGENTS.md
    • ACPClient/AGENTS.md
    • AppServerClient/AGENTS.md
    • AgmenteTests/AGENTS.md
    • AgmenteUITests/AGENTS.md
    • e2e/AGENTS.md

Local Path Config

  • Use root .agmente.paths for machine-local external repo paths.
  • Required vars: AGMENTE_ACP_REPO, AGMENTE_CODEX_REPO.
  • Read this file when running upstream-spec/source checks.

Core Architecture

  • Agmente supports two protocol modes: ACP and Codex app-server.
  • Protocol is detected after initialize and routed to protocol-specific view models.
  • Keep ACP and Codex behaviors separate unless a shared abstraction is intentional and tested.

Contribution Rules

  • Prefer component-local changes; avoid cross-cutting refactors unless needed.
  • Do not add personal local filesystem paths to tracked files.
  • If behavior differs by protocol, document both expected paths.
  • For UI changes, keep accessibility identifiers stable or update UI tests in the same PR.

Documentation Rules

  • Update Agents.md and this file in the same PR when:
    • adding a new top-level component,
    • adding protocol support,
    • changing cross-component architecture.
  • Update nested <component>/AGENTS.md in the same PR when behavior in that component changes.
  • If no docs are needed, include Docs: N/A in PR notes.

Testing Expectations

  • Run focused tests for touched components first.
  • For protocol lifecycle changes, run both:
    • AgmenteTests/CodexServerViewModelTests
    • AgmenteTests/ViewModelSyncTests
  • For UI flow or accessibility changes, run relevant AgmenteUITests.
  • Codex E2E remains opt-in via environment configuration (see AgmenteUITests/AGENTS.md).
  • For manual or simulator E2E requests, resolve the scenario from e2e/scenarios/ before consulting any skill.
  • Treat run, execute, validate, and reproduce E2E requests as execute-only and agentic: run the real backend and tools, capture evidence, evaluate assertions, and clean up.
  • Do not modify repo files during an execute-only E2E run unless the user explicitly asks to fix, stabilize, add hooks, or make the scenario executable.
  • For simulator interaction, prefer the xcodebuildmcp CLI as the source of truth for available features:
npx -y xcodebuildmcp@latest tools
  • Do not assume the in-session mcp_xcodebuildmcp_* subset includes interaction tools (tap, swipe, gesture, type-text, long-press).
  • If any required UI automation tool is missing from MCP exposure, switch to xcodebuildmcp CLI for simulator interaction.

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.