Privacy notice
Privacy
GovBot Environmental Site Context is a read-only research service. You do not need to give it personal information.
In brief: the service keeps no application database of conversations or request bodies. It keeps bounded in-memory caches of agency source records, and the hosting platform keeps operational request logs for 30 days.
What the service receives
Tool requests contain public research inputs: coordinates and search radii for a public research point, or a public 12-digit EPA Registry ID. The service does not keep an application database of conversations or request bodies.
No sign-in, account, API key, payment, subscription or file upload is needed. You do not need to provide your name, contact details or a personal home address. Please use public research points rather than personal addresses.
Where requests go
To answer a request, the service sends those public inputs to fixed public agency services:
- U.S. Fish and Wildlife Service National Wetlands Inventory maps, hosted by the U.S. Geological Survey Wetlands and Wildlife Information Management (WIM) program.
- U.S. Fish and Wildlife Service Ecological Services critical habitat service.
- U.S. Environmental Protection Agency ECHO facility search and Detailed Facility Report.
The service uses no source account or API key with these agencies. Each agency has its own practices for requests it receives.
In-memory caching
To limit repeat calls to agency services, the service holds agency source responses, and the query keys that contain the public research inputs for each lookup (point coordinates and search radius, or an EPA Registry ID), in bounded process memory. The time limits below govern eligibility for reuse after retrieval:
- Ordinary source responses: 10 minutes.
- Agency layer metadata: 24 hours.
A reuse time limit is not a deletion deadline. An expired entry may remain in memory until a later lookup of the same key, replacement or capacity eviction, or process exit. A cache timestamp shows when a record was retrieved from the source. It is not the date or vintage of the agency dataset.
Operational logs
The service runs on Google Cloud. The hosting platform keeps operational HTTP request logs, which can include the IP address a request came from, the request path, the time of the request, the response status, latency and user agent. These logs are retained for 30 days under the project log retention setting.
Support email
If you write to hello@govbot.ai, your message is kept until it is manually deleted. There is no fixed deletion schedule.
OpenAI
If you use this service through ChatGPT, OpenAI handles your conversations under its own, separate privacy practices.
Contact
Privacy questions: hello@govbot.ai. See also the terms of use and the product guide.