Bring public wildfire evidence into your AI workflow.
Three read-only tools return French Meadows project evidence, reference wildfire-risk summaries and a source-linked review.
Connection details
| Connector endpoint | https://wildfire.govbot.ai/mcp |
|---|---|
| Transport | Streamable HTTP |
| Authentication | None for the public pilot; no API key or account |
| Coverage | French Meadows: county mapped project area, Unit 43 and Unit 125 |
| Support | hello@govbot.ai |
Custom connector availability depends on your AI provider, plan and organization settings. A custom connection is not a public directory listing.
Tools
| Tool | What it returns |
|---|---|
wildfire_get_project | The real public project, approved area IDs, treatment counts and source links. |
wildfire_estimate_reference_risk | Reference annual burn probability, expected annual exposed area, analyzed acreage, coverage and method for one approved area. |
wildfire_export_review | Markdown review text with evidence, reference risk, sources and limits. It does not save or publish a file. |
The area IDs are county-project, treatment:317 (Unit 43) and treatment:338 (Unit 125). Arbitrary map drawing is available separately in the web app; the connector does not accept private coordinates or notes.
Try these questions
- What real project does GovBot Wildfire cover, and which areas can I assess?
- Estimate the reference annual wildfire likelihood for the whole French Meadows project.
- Compare reference wildfire risk at county treatment Units 43 and 125.
- Give me a sourced Markdown review of French Meadows with the reference risk results.
Read the result correctly
The WRC 2024 edition uses end-of-2020 landscape conditions for burn probability: a 270 m underlying model distributed on a 30 m grid. It is regional baseline screening, not a current forecast or an individual-home assessment.
The mean probability is an average across locations, not the chance of any fire somewhere in the project. Comparing units does not estimate the effect of their treatments. No tool calculates property loss, BCA, ROI or a FEMA-approved result.
Public source evidence stays separate from published insurance findings. Read about the sources and method.
Data handling
The service processes approved public identifiers and returns retained evidence. It creates no user account and makes no generative-AI or Jev calls. Your assistant provider handles your conversation under its own terms.