🐦 WingIt-MCP

An MCP server that helps birders wing it — generating personalized target checklists for new lifers.

WingIt-MCP is a Go-based Model Context Protocol server that combines your personal eBird checklists with recent community sightings to recommend nearby species you’ve never seen — your next lifers — in a ready-to-use field checklist.

WingIt-MCP is written in idiomatic Go and implements the Model Context Protocol (MCP) to make its tools and resources discoverable to AI hosts like Claude Desktop.

Use

WingIt-MCP needs your personal checklist and a source of recent observations.

Required environment variables

• WINGIT_PERSONAL_JSON: path to your personal eBird checklist export (JSON).

Recent observations

You can run in one of two modes:

• Offline: set WINGIT_RECENT_JSON to a JSON file of recent observations.
• Live: set WINGIT_EBIRD_TOKEN to your eBird API token and pass location as "lat,lon".

Examples:

• location="42.47,-76.45"
• location=" 42.47 , -76.45 "

Latitude must be between -90 and 90, and longitude between -180 and 180.

Roadmap

v0.2.0 – Live eBird data + MCP polish

Planned improvements for the next release:

• Live eBird API support for recent nearby sightings (env-gated)
• Simple caching to reduce API calls and improve demo repeatability
• Clear runtime mode selection:
  • offline fixtures via WINGIT_RECENT_JSON
  • live API via WINGIT_EBIRD_TOKEN
• MCP integration smoke test (tools / prompts / resources)
• Friendlier output formatting (frequency bands: high / medium / low)

Track progress in the GitHub milestone: v0.2.0 – Live eBird data + MCP polish.