Feed America launches Model Context Protocol (MCP) server for AI assistant food-help integration
Feed America (EIN 92-1761881) launched a Model Context Protocol (MCP) server in Q2 2025, providing 7 native AI tools for food-help discovery via Anthropic Claude, OpenAI custom GPTs, and other MCP-compatible assistants. Tools include resource search by ZIP, geocoding, eligibility checking, urgent-need filtering. Public OpenAPI 3.0 spec also published.
Feed America (EIN 92-1761881) announced the public launch of a Model Context Protocol (MCP) server for AI assistant food-help discovery. The MCP server, available at /mcp/v1, provides 7 native AI tools that Anthropic Claude, OpenAI custom GPTs, and other MCP-compatible AI assistants can call directly to surface food-help information for users.
The 7 MCP tools
- search_resources — ZIP-based search for food pantries, food banks, soup kitchens, school meals, summer meals, WIC, FQHC, SNAP retailers
- nearby_resources — lat/lng-based search with distance filtering
- resource_counts — per-mode result counts within 150mi of a ZIP
- urgent_resources — open-now filter for someone hungry right now
- geocode_address — address-to-lat/lng for follow-up searches
- snap_eligibility — household-size + income eligibility check
- disaster_info — FEMA active declarations + D-SNAP eligibility for a location
Why MCP matters
MCP is Anthropic's open standard for AI tool integration. By implementing MCP natively, Feed America's data becomes directly accessible to AI assistants without per-platform custom integrations. A user asking Claude or any MCP-compatible assistant "where can I find food help near 90210?" can have the assistant call our tool directly and surface accurate, real-time food-help data.
Public OpenAPI 3.0 specification also published at /api/openapi.json for OpenAI custom GPT integration. Discovery manifest at /.well-known/mcp.json.