Traffic Camera Hub

For AI assistants

Give Claude eyes on the road.

Traffic Camera Hub is available as a remote Model Context Protocol connector. Add it to Claude on the web, desktop or mobile and it can pull live camera images while it answers — say, while a maps connector plans your route.

Connector URL

https://trafficcams.bradylawson.com/api/mcp

Add it to Claude

  1. 01

    Open connector settings

    In claude.ai go to Settings → Connectors → Add custom connector. The same menu is in the Claude desktop and mobile apps.

  2. 02

    Paste the URL

    Use the connector URL above. No sign-in or API key is required.

  3. 03

    Ask about the road

    Start a new chat and try “What does traffic look like near I-66 between DC and Fairfax right now?” Claude finds the relevant cameras and shows the images inline.

Tools the connector exposes

find_cameras_near
Find cameras within 1–50 km of a point, sorted by distance, with route, milepost and a public image link. Hosts that support MCP Apps also get an embedded map with clickable pins.
get_camera_image
Fetch the current still image from one camera so the model can look at the road itself. For video-only cameras, a current frame is taken from the live stream.
view_camera
Open the interactive viewer for one camera: live video where the camera streams, an auto-refreshing still where it doesn’t — the same playback as this site.

Coverage & limits

  • Cameras from ~34 US state transportation departments, with new cameras picked up daily.
  • Live video and auto-refreshing stills via MCP Apps (spec 2026-01-26). Hosts without MCP Apps get text and image responses instead.
  • Heavy automated use may be rate-limited and receive HTTP 429 responses.