# ATL Transit Live

Unofficial MARTA service reliability, realtime rail, generated delay alerts, and developer API feeds.

## Primary Docs

- Developer API overview: https://atltransit.live/developers
- API docs page: https://atltransit.live/api-docs
- OpenAPI schema: https://atltransit.live/api/v1/openapi.yaml
- Full LLM context: https://atltransit.live/llms-full.txt

## Public API

This list is the full set of documented operations. `GET /api/v1/openapi.yaml` is the
authoritative schema; prefer it over this summary when they disagree.

### Account and metadata

- `GET /api/v1/health`: verify an API key and read its resolved plan limits.
- `GET /api/v1/openapi.yaml`: fetch the OpenAPI document.
- `GET /api/v1/billing/plans`: list the plan catalogue and each plan's hard limits.

### Alerts

- `GET /api/v1/alerts`: service alerts across every mode.
- `GET /api/v1/rail/alerts`: stored rail alerts as JSON or SSE.
- `GET /api/v1/bus/alerts`: stored bus alerts as JSON or SSE.
- `GET /api/v1/streetcar/alerts`: stored streetcar alerts as JSON or SSE.
- `GET /api/v1/rail/alert-stats`: per-line or per-station rail alert statistics.

### Rail

- `GET /api/v1/rail`: rail line metadata and links to rail feeds.
- `GET /api/v1/rail/realtime`: normalized MARTA rail realtime arrivals as JSON or SSE.
- `GET /api/v1/rail/positions`: current rail snapshot, one row per train-at-station-direction.
- `GET /api/v1/rail/on-time`: rail on-time performance.
- `GET /api/v1/rail/headways`: inferred rail headways (time between trains).
- `GET /api/v1/rail/stations`: rail stations and the lines serving them.
- `GET /api/v1/rail/{line}/reliability`: rail line reliability rollups as JSON or SSE.
- `GET /api/v1/rail/{line}/trips/today`: a rail line's trips for a service day.

### Bus

- `GET /api/v1/bus`: bus route metadata and reliability snapshots.
- `GET /api/v1/bus/realtime`: GTFS-RT bus vehicle positions as JSON or SSE.
- `GET /api/v1/bus/{route_id}/realtime`: GTFS-RT bus positions for one route as JSON or SSE.
- `GET /api/v1/bus/speed`: average bus speed.
- `GET /api/v1/bus/service-delivery`: observed versus scheduled buses.
- `GET /api/v1/bus/{route_id}/reliability`: bus route reliability rollups as JSON or SSE.
- `GET /api/v1/bus/{route_id}/trips/today`: a bus route's trips for a service day.

### Stops

- `GET /api/v1/routes/{route_id}/stops`: current-feed stop geometry and current MARTA marker predictions for one route.
- `GET /api/v1/stops/{stop_id}/predictions`: current MARTA and ATL Transit Live predictions for a stop in required route context.
- `GET /api/v1/stops/{stop_id}/prediction-history`: bounded 7-, 30-, or 90-day aggregate prediction analytics (Max only; no raw observations).

### Streetcar

- `GET /api/v1/streetcar`: streetcar route metadata.
- `GET /api/v1/streetcar/realtime`: GTFS-RT streetcar vehicle positions as JSON or SSE.
- `GET /api/v1/streetcar/{route_id}/realtime`: GTFS-RT streetcar positions for one route as JSON or SSE.
- `GET /api/v1/streetcar/{route_id}/reliability`: streetcar reliability rollups as JSON or SSE.
- `GET /api/v1/streetcar/{route_id}/trips/today`: a streetcar route's trips for a service day.

### System-wide

- `GET /api/v1/system/reliability`: system-wide reliability.
- `GET /api/v1/system/today`: today's system-wide cancellation count.

### Rider feedback

- `GET /api/v1/lines/{id}/feedback`: rider ratings and approved comments for a line.
- `POST /api/v1/lines/{id}/feedback`: submit a rider rating, and optionally a comment.

### Streaming, webhooks, and bulk exports

- `GET /api/v1/events/stream`: persistent SSE event stream.
- `POST /api/v1/webhooks/endpoints`: create a webhook endpoint.
- `GET /api/v1/webhooks/endpoints`: list webhook endpoints.
- `PATCH /api/v1/webhooks/endpoints/{id}`: update an endpoint, optionally rotating its secret.
- `DELETE /api/v1/webhooks/endpoints/{id}`: delete (disable) a webhook endpoint.
- `GET /api/v1/webhooks/deliveries`: webhook delivery attempts.
- `GET /api/v1/history/exports`: historical export manifests (Max only).
- `POST /api/v1/history/exports`: enqueue generation of a bulk export (Max only).
- `GET /api/v1/history/exports/{id}/download`: short-lived signed download URL for an export.

## Notes For LLM Clients

- REST endpoints return normalized ATL Transit Live response envelopes with `data_sources`, cursor pagination where applicable, and computed delay/reliability fields.
- Source alert structures preserve array-valued `impact_type`, `affected_ids`, and `microblog_message` fields from the LLM JSON schemas.
- Refreshed resources, including alerts, realtime positions, and reliability rollups, return JSON by default and SSE when requested with `Accept: text/event-stream` or `format=sse`.
- Alert resources read stored PostgreSQL `generated_service_alerts` rows when API persistence is configured and support `status`, `route_or_line`, `station`, `source`, `limit`, and `cursor` filters.
- Successful REST calls are metered as `api_request`.
- SSE output uses `sse_connection_minute` and `sse_event`.
- Webhook events are signed with HMAC-SHA256 using `Marta-Fail-Signature`.
- Current plan limits are a snapshot; read `GET /api/v1/billing/plans` for the effective catalogue.
- Free: 20 requests/5 minutes, 300/day, max `limit=20`, 1 API key, no SSE, webhooks, or bulk exports.
- Regular: 600 requests/5 minutes, 50,000/day, max `limit=100`, 2 SSE connections, 5 webhooks, 5 API keys, no bulk exports.
- Max: 6,000 requests/5 minutes, 1,000,000/day, max `limit=1,000`, 25 SSE connections, 25 webhooks, 25 API keys, bulk exports enabled.
- Rail realtime responses include `rail_status.lines[].stations[].trains[]`, `line_delay_summary`, `station_delay_summary`, and cursor pagination.