agentleFS
satoshinji2992satoshinji2992/autokeeper/AGENTS.md

autokeeper

AGENTS.md3 starsChanged 24 days agoMITView on GitHub

AutoKeeper runs Chinese Call of Cthulhu 7th-edition games on Python 3.12+. Prerequisites: Python 3.12+, uv, Node.js 18+. Use four-space indentation, annotate new interfaces, and use snake_case functions/modules, PascalCase classes, and UPPER_CASE constants. Preserve Chinese UI terminology and the colored ASCII aesthetic. Retain original module images and readable text; unlock spoiler images with their clues. Ruff targets 120-character lines and checks imports, modernization, and bugs; its configuration permits Chinese punctuation. Keep Textual styles in .tcss files. Write concise, professional Chinese UI…

  • Reads credentials
  • 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 repository4 other files it gives its agents.

The file

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

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

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈0

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 3

    Very specific

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

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

    Use pytest and pytest-asyncio (automatic async mode).

    line 35
  2. Very specific:

    Use focused commits with concise Chinese messages, optionally prefixed by an area such as 地图:.

    line 39
  3. Very specific:

    Keep API keys in environment variables or ignored .env files.

    line 47

≈ “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

Repository Guidelines

Project Structure & Module Organization

AutoKeeper runs Chinese Call of Cthulhu 7th-edition games on Python 3.12+.

  • src/autokeeper/: engine/ manages turns/state; agents/ and harness/ handle models; rules/ implements mechanics; plugins/ runs tested extensions; net/ supports multiplayer; module/ handles scenarios (module/package.py is the on-disk layout of one); world/ is the persistent world record. paths.py owns every location on disk: ask it instead of joining paths yourself.
  • src/autokeeper/tui/: app.py (the Textual app) and library.py (save index, module loading, art_text); screens/ full pages (title, setup, game, cards, LAN host, lobby, saves, import); modals/ dialogs over a screen; widgets/ panels, speaker, pages, tips; drawing/ title art, cover and dice. Each .tcss sits beside its code.
  • bridge/: Node.js JSON-lines bridge to BCDice.
  • tests/: tests, conftest.py fixtures (every test gets its own paths.DATA, paths.DEV_OUTPUT and paths.USER_DIR), and cassettes/ recordings.
  • modules/<id>/: the three bundled scenarios: module.yaml plus places.yaml, people.yaml, clues.yaml, story.yaml, situation.yaml, plugins/, and art/{places,people,items,ascii}/ with one art/credits.yaml. src/autokeeper/module/example/ is the format fixture. Never commit derived files into a module (prepared pictures, compiled skills); they live under data/cache/ and skills/module-*.
  • skills/: shared Agent Skills the engine injects into prompts (coc7-core, keeper-craft, pingshu-voice, ascii-art; read through skills.py and paths.skill). Shared skills are tracked; skills/module-* are compiled and ignored. docs/ contains technical documentation; docs/img/ holds screenshots.
  • dev/: developer scripts (README screenshots, public-share tunnel); their output and eval/autoplay reports go to ignored dev/output/.

Build, Test, and Development Commands

Prerequisites: Python 3.12+, uv, Node.js 18+.

  • uv sync --extra claude: install dependencies.
  • npm ci --prefix bridge --no-audit --no-fund: install the locked dice-bridge dependencies.
  • uv run autokeeper doctor: inspect the local environment.
  • uv run autokeeper: launch the terminal game; uv run autokeeper serve starts browser play.
  • uv build: build Python distributions with Hatchling.
  • uv run ruff check src tests: run lint and import-order checks.
  • uv run pytest -q: run the test suite.
  • uv run autokeeper check src/autokeeper/module/example: validate the format example.
  • uv run agentskills validate skills/<name>: validate a skill.

Coding Style & Naming Conventions

Use four-space indentation, annotate new interfaces, and use snake_case functions/modules, PascalCase classes, and UPPER_CASE constants. Preserve Chinese UI terminology and the colored ASCII aesthetic. Retain original module images and readable text; unlock spoiler images with their clues. Ruff targets 120-character lines and checks imports, modernization, and bugs; its configuration permits Chinese punctuation. Keep Textual styles in .tcss files. Write concise, professional Chinese UI copy: state the action, status or outcome directly; avoid chatty phrasing, personified loading messages, redundant instructions and generic requests to rephrase after technical errors. Put detailed rules in help, and avoid internal agent names, marketing claims, or tactical coaching. Preserve the narrative voice in story text and NPC dialogue.

Testing Guidelines

Use pytest and pytest-asyncio (automatic async mode). Name files test_<area>.py and functions test_<behavior>. Use scripted agents, FakeDice, recordings, and temporary directories. BCDice tests skip without its Node bridge. No numeric coverage threshold; add regression tests for behavior changes.

Commit & Pull Request Guidelines

Use focused commits with concise Chinese messages, optionally prefixed by an area such as 地图:. PRs must describe behavior changes, linked issues, and validation. Include screenshots for UI/artwork changes. Document changes to commands, configuration, and module formats.

Rapid Iteration

Backward compatibility with older saves, configuration, network protocols, and internal interfaces is not required at this stage. Change formats directly and remove superseded adapters rather than maintaining parallel implementations. Verify current-format save/load, interrupted-action recovery, reconnection, and multiplayer consistency.

Configuration & Local Data

Keep API keys in environment variables or ignored .env files. User configuration lives in ~/.config/autokeeper/config.toml. Keep game data, imported copyrighted material, and generated module skills in their ignored locations: data/ (layout in paths.py), modules/local/, modules/inbox/, every module's source/, dev/output/, and skills/module-*.

Disk saves share repeated log bodies. Write through engine.save.dump_save and read through engine.state.read_save / check_save; direct GameState.model_validate_json is only for ordinary runtime snapshots, not save files. Preserve event IDs, audience, order and recovery receipts when changing storage.

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.