sl-trafiklab-api（SL trafiklab API 搜索）

SL Trafiklab API Skill

This skill retrieves real-time departure and deviation information from SL (Stockholm public transport). It does NOT require registration of stops or routes to make queries—you can query any stop or route on demand. The favourite_stops and favourite_routes configuration is only for autonomous background monitoring (e.g., heartbeat checks, cron jobs that proactively notify about disruptions).

The most critical task when handling background routes is to filter out noise. The user should not be notified about disruptions at stops/lines they're not actually using on that specific route.

State Storage

Preferences for autonomous monitoring are maintained in your workspace at .sl/preferences.json. Initialize this file with empty arrays if it does not exist. See references/api.md for the exact JSON format, which includes favourite_stops and favourite_routes.

To save updates to the user's monitoring preferences, simply write valid JSON to .sl/preferences.json.

Core Actions

For all specific API endpoints (curl commands), strict payload formats, and data extraction instructions, you must refer to references/api.md.

1. search_sl_sites: Finds the numeric Site ID for a given stop name. The Departures and Deviations APIs require numeric IDs. Warning: Large payload, see references/api.md for safe filtering via jq.
2. fetch_departures: Fetch real-time departures from ANY site (does not need to be in favourites). Returns upcoming departures with scheduled/expected times, destinations, and deviations. Optional filters: line, transport_mode, direction, forecast window.
3. fetch_deviations: Fetch disruption info for ANY site/line (does not need to be in favourites). You must utilize the strict assessment logic defined in references/api.md when filtering for autonomous notifications.
4. manage_preferences: Overwrite .sl/preferences.json to configure favourite_stops and favourite_routes for autonomous monitoring only.

Autonomous Directives (Background Monitoring)

During autonomous execution (e.g., background heartbeat or cron job):

1. Load .sl/preferences.json.
2. Check favourite_stops:

- For each stop, optionally filter by its registered lines and/or transport_modes when fetching departures - Fetch deviations affecting these stops

1. Check favourite_routes:

- Execute detailed deviation checking using the strict assessment logic from references/api.md - For each leg of the route, verify disruptions actually affect the user's specific boarding/alighting stops

1. Compare returned deviation IDs against context memory.
2. Only send a notification if a new, relevant disruption is detected.
3. Adhere to Trafiklab's limit of maximum 1 request per minute.