AGENT SKILL

open-wearables

Operating router for the Open Wearables FastAPI backend, wearable provider integrations, React portal, and FastMCP server.

License 许可
MIT
最近更新
2026-09-03
Stars
★ 351
Forks
27

01 前言 Front Matter

name
open-wearables
license
MIT
metadata
{ "disco-role": "operating" }
description
Operating router for the Open Wearables FastAPI backend, wearable provider integrations, React portal, and FastMCP server.
disable-model-invocation
true

02 资源文件 6

03 SKILL.md 正文

Open Wearables

Use this repo skill for Open Wearables, a monorepo that aggregates wearable and health data through a Python/FastAPI backend, React/TanStack Start portal, and FastMCP server for AI assistants.

Route the task

  • Backend API, auth/developer credentials, users, connections, summaries, events, timeseries, sync status, seed data, outgoing webhooks, service/repository/model/schema work, migrations, DB/Redis/Celery/Svix/AWS config, and backend tests → backend-core.
  • Wearable provider strategies, provider enum/factory registration, OAuth handlers, workout/24-7 data handlers, incoming provider webhooks, coverage declarations, coverage docs, provider sync/backfill, and provider-specific tests → provider-integrations.
  • React portal routes, protected auth layout, dashboard/users/settings/coverage/sync/webhooks pages, API hooks/services/query keys, runtime VITE_API_URL, TanStack Router/Query, components, styling, and frontend test/lint/build work → frontend-portal.
  • FastMCP server startup, assistant configuration, MCP tools, REST client behavior, prompts, mocked MCP tests, and MCP API-key/backend failure handling → mcp-server.

Read shared references

  • setup.md — prerequisites, install commands, service topology, and safe validation commands.
  • architecture.md — monorepo package map, backend/provider/frontend/MCP data flow, and cross-skill ownership.
  • troubleshooting.md — shared install/import/service/config/credential failures and where to route deeper debugging.
  • repo-provenance.md — source snapshot for staleness checks before refreshing this skill.

Run the safe root checker when a checkout is available:

python skills/disco/open-wearables/scripts/check_open_wearables_install.py --repo-root .

This checker is read-only. It verifies package metadata, expected directories, docs navigation presence, and local tool availability; it does not start services, install packages, call provider APIs, or read real secrets.

Package and service baseline

  • Backend package: open-wearables==0.7.0, Python >=3.13, FastAPI, SQLAlchemy 2.0, PostgreSQL, Redis, Celery, Svix, AWS/S3 optional storage, Ruff, Ty, Pytest.
  • Frontend package: frontend-app==0.7.0, Node >=22, pnpm >=10, React 19, TanStack Start/Router/Query, Tailwind CSS 4, shadcn/ui, Vitest, oxlint, Prettier.
  • MCP package: open-wearables-mcp==0.1.0, Python >=3.13, FastMCP, httpx, Pydantic settings, start = app.main:main.
  • Local stack: Docker Compose starts backend, frontend, PostgreSQL, Redis, Celery workers/beat/Flower, and Svix services. Use a disposable dev/test stack for seed/reset/replay utilities.

Boundary rules

  • Do not commit real provider credentials, API keys, JWTs, OAuth tokens, Svix secrets, AWS keys, master encryption keys, or .env files.
  • Keep backend endpoint changes synchronized with frontend/MCP callers and docs. External API endpoints tagged External: * require API Reference navigation updates.
  • Provider coverage is the source of truth for /api/v1/meta/coverage and the portal coverage matrix; changes belong in provider-integrations even when surfaced in backend routes or frontend UI.
  • MCP is REST-client-driven. Do not bypass the backend by importing backend database models or direct DB access into the MCP package.
  • Runtime frontend API URL must go through resolveApiUrl() / API_CONFIG.baseUrl; do not inline import.meta.env.VITE_API_URL in application code.
  • Prefer focused native checks after edits: backend targeted pytest, provider coverage/import tests, frontend Vitest/lint/build when dependencies are installed, and MCP mocked pytest.

Staleness check

Before relying on this skill for a different checkout, read repo-provenance.md. If the commit, package versions, route/plugin entry points, provider inventory, frontend scripts, or MCP tool names changed, use refresh-repo-skill rather than patching this skill by hand.

04 同仓库更多 Skills

VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-triton-opt

Make a single PassNet Triton kernel fast AND numerically faithful: performance model (when a replacement can win at all), block/grid/warp tuning, autotune policy, launch overhead, and per-op numeric recipes to pass the dtype baseline tolerances. Use when a pass matches and is correct but speedup below expected, or when correctness fails by small numeric margins.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-solve

END-TO-END playbook for solving one PassNet sample: analyze the computation graphs, decide the optimization strategy (what to fuse, what to replace, what to leave alone), drive the iteration loop, and maximize the sample score. This is the ENTRY skill — invoke it first for any PassNet optimization task; it tells you when to use passnet-pattern-fusion, passnet-triton-opt and passnet-feedback.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-skill

PassNet GPU kernel optimization via compiler passes. Design and implement Triton-based optimization passes, create pass files under ./pass_dir/, self-evaluate with pass_evaluator, and iterate to maximize GPU speedup.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-pattern-fusion

Author PassNet pass files: write patterns that actually MATCH the FX graph, pick fusion regions (which ops to absorb into one Triton kernel), and structure multi-pass submissions with the mandatory shared-dispatch architecture. Use when creating or fixing pattern()/replacement_args()/replacement_func() files, when a pass "failed to match", or when deciding how to fuse multiple kernels/ops.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-orchestrate

Round/batch planning and decision support for PassNet graph optimization. Use for multi-sample triage, worker allocation, eval-budget policy, and post-eval keep/revert decisions. For solving one concrete sample end to end, use passnet-solve as the entry skill; consult this skill only when you need broader planning or a specific decision gate.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-feedback

Fast, deterministic feedback for PassNet work: pre-flight pattern/match verification WITHOUT burning a GPU evaluation, per-node bottleneck analysis of a sample's graphs, and eval-log parsing into per-variant status + estimated score + failure classification. Use BEFORE every GPU evaluation (check_pattern), at the START of a sample (analyze_graph), and AFTER every evaluation (parse_eval_log).

★ 351