About FireWatch
FireWatch shows active wildfire incidents across the United States. Data is sourced from the National Interagency Fire Center's WFIGS feeds and refreshed every 15 minutes. It is designed to aggregate and deliver a clean MCP server of information to LLMs.
Not affiliated with Watch Duty. Not an emergency service — always follow official guidance from local authorities.
For Agents
LLM Info (llms.txt)
Structured summary of FireWatch for large language models — endpoints, data sources, and usage guidance.
MCP Server
Model Context Protocol endpoint so AI agents (Claude, ChatGPT, Cursor) can query live wildfire data as tools.
Building an agent on top of FireWatch? Point it at the MCP endpoint and use the
search_firesandget_firetools.