docs: extract plugin contracts into modular docs and slim CLAUDE.md

CLAUDE.md had grown to 545 lines / 50 KB — most of it plugin-author
content that other AI tools (Cursor, Copilot, Codex, Aider) and humans
without AI never reach. Extract the plugin surface into 10 focused
docs and a JSON Schema for plugin.json, then slim CLAUDE.md to a
156-line navigable index.

New docs (~999 lines total, all self-contained):

  docs/PLUGIN_AUTHORING.md            — entry point and quickstart
  docs/plugin-manifest.md             — plugin.json field reference
  docs/plugin-visualization-contracts.md — setRenderer / overlay / note-state
  docs/plugin-audio-mixer.md          — fader registration
  docs/plugin-logging.md              — context["log"] + env vars
  docs/plugin-diagnostics.md          — server_files / callable
  docs/plugin-keyboard-shortcuts.md   — registerShortcut + scopes
  docs/plugin-sibling-imports.md      — load_sibling pattern
  docs/websocket-protocol.md          — /ws/highway message reference
  docs/testing-plugins.md             — pytest fixtures + Playwright

  schema/plugin.schema.json           — Draft 2020-12 schema for
                                        plugin.json; license enum
                                        mirrors CONTRIBUTING's curated
                                        allowlist. Backs CI validation
                                        and the plugin-validate skill.

CLAUDE.md slim (581 lines changed, -485):

  - Removed ~300 lines of plugin-author prose (now in docs/).
  - Kept architecture quick reference, running the app, testing,
    git workflow, versioning, song formats, frontend/backend
    conventions, plugin authoring INDEX (table → docs/), first-hour
    pitfalls, "For AI agents" footer.
  - Anchor stubs preserved next to the new index entries so deep
    links from specs/001-slopsmith-platform/analyze.md still resolve.

Verification:
  python -c "import json,glob,jsonschema; s=json.load(open('schema/plugin.schema.json')); [jsonschema.validate(json.load(open(p)), s) for p in sorted(glob.glob('plugins/*/plugin.json'))]"
  # ok — validates highway_3d, app_tour_library, app_tour_settings
Signed-off-by: Miguel_LZPF <mgcdreamer@gmail.com>
This commit is contained in:
Miguel_LZPF
2026-06-18 00:38:50 -07:00
committed by Bret Mogilefsky
parent cad78857fb
commit 1214a6c0a0
12 changed files with 1013 additions and 625 deletions
+48
View File
@@ -0,0 +1,48 @@
# Plugin sibling imports — `load_sibling`
When a backend plugin spans multiple Python files, **use `context["load_sibling"]`** rather than bare `import` statements. This is the slopsmith#33 contract.
## The problem
The plugin loader inserts each plugin's directory onto `sys.path` so `from extractor import X` works. But Python caches imports by **module name** in `sys.modules`. Two plugins that each ship a top-level `extractor.py` (or any other generic name — `util.py`, `client.py`, `parser.py`, `config.py`, …) collide: whichever loads first wins, and the other plugin's `from extractor import X` either gets the wrong module or fails with `cannot import name 'X' from 'extractor'`.
## The fix
`context["load_sibling"](name)` loads the sibling under a namespaced module name `plugin_<id>.<name>`, so each plugin gets its own copy. The `<id>` portion is bijectively encoded so reverse-DNS-style ids like `com.example.foo` work without colliding: `_``_5f_`, `.``_2e_`.
```python
def setup(app, context):
extractor = context["load_sibling"]("extractor")
PsarcReader = extractor.PsarcReader
# …
```
## Notes
- **`name` is a bare module name** — no `.py` suffix, no slashes, no `.`. The helper raises `ValueError` for path traversal / format issues and `ImportError` for missing files.
- **Both single-file siblings (`extractor.py`) and package-form siblings (`extractor/__init__.py`) work.** Package form wins when both exist (matches CPython's import-resolution precedence).
- **Relative imports between siblings work** — `from .shared import X` in a top-level helper, `from ..shared import X` from inside a sibling package. The synthetic parent package `plugin_<id>` carries the plugin directory in its `__path__`.
- **`from . import sibling` (attribute-style) also resolves**: loaded children are exposed as attributes on the parent package.
- **Repeat calls return the cached module.** Concurrent first-time calls are serialized via per-module locks so no caller observes a half-initialized module.
- **Don't mix `load_sibling` and bare `import` for the same module** — they'd execute the file twice and split module-level state.
## Migration
Bare `import sibling` from `routes.py` still works during the transition period, but the loader prints a startup warning when it detects two plugins shipping a same-named top-level module — covering both `.py` files and package directories. Migrate to `load_sibling` to silence the warning and immunize your plugin from future ecosystem collisions.
## Verification
The collision-detection logic is encoded in `tests/test_plugins.py`. Run:
```bash
pytest tests/test_plugins.py -v
```
If you're authoring a plugin with a top-level helper, add a test that imports your plugin under a `reset_plugin_state` fixture to confirm clean import behaviour. See [testing-plugins.md](testing-plugins.md).
## Related
- [PLUGIN_AUTHORING.md](PLUGIN_AUTHORING.md) — guide index
- [plugin-manifest.md](plugin-manifest.md) — declaring `routes.py`
- [plugin-logging.md](plugin-logging.md) — `context["log"]` (also exposed via `context`)
- [testing-plugins.md](testing-plugins.md) — `reset_plugin_state` fixture