Five tools covering the USDA FoodData Central workflow — from discovery to detailed nutrient analysis:

Search USDA FoodData Central foods by keyword, UPC/GTIN code, or ingredient.

• Covers all FDC data sources: SR Legacy (common whole foods, complete nutrient profiles), Foundation, Survey FNDDS, and Branded (packaged products)
• Defaults to SR Legacy; include "Branded" in dataType for packaged products or UPC lookup
• Brand owner filter to narrow branded results (e.g. "General Mills")
• Food category filter (e.g. "Poultry Products", "Vegetables and Vegetable Products")
• Pagination via pageSize (up to 50) and pageNumber
• Returns FDC IDs and a preview of key nutrients (energy, protein, fat, carbs) — use usda_get_food for the full profile

Full nutrient profile for one food by FDC ID.

• All available nutrients (or a filtered subset via nutrients[]) with amounts per 100g
• Optional portion scaling — provide quantity + unit to scale values (e.g. quantity=200, unit="g" → per-200g values)
• unit="serving" scales to the food's first defined portion weight
• Returns all named portion definitions (allPortions[]) alongside the serving info
• Filtering nutrients[] to specific IDs strongly reduces context size for common queries (use usda_list_nutrients to look up IDs)

Batch nutrient fetch for 2–20 FDC IDs.

• All values per 100g (no portion scaling in batch mode)
• nutrients[] filter strongly recommended — full profiles for 20 foods are large
• Per-item partial failure: failed[] carries IDs that returned no data, so one missing food doesn't abort the batch
• More efficient than N individual usda_get_food calls when you already have FDC IDs

Side-by-side nutrient comparison for 2–5 foods.

• Returns a pivot table — one row per nutrient, one column per food — and formats it as a markdown table
• Defaults to the 12 most commonly compared nutrients (energy, protein, fat, saturated fat, carbs, fiber, sugars, sodium, potassium, calcium, iron, vitamin C)
• Pass custom nutrients[] for specific comparisons (e.g. just iron and vitamin C)
• All values scaled to a common gram basis (default 100g; override with quantity + unit)
• Proceeds with valid foods when some IDs aren't found — only fails when fewer than 2 IDs return data
• Rows where all values are null (nutrient not tracked for any of the selected foods) are filtered out

FDC nutrient reference table — all ~150 tracked nutrients.

• Returns IDs, names, SR reference numbers, units, and categories
• Optional category filter: macronutrients, vitamins, minerals, lipids, amino_acids, other
• Resolves nutrient names (e.g. "vitamin C") to FDC IDs (1162) for use in nutrients[] params
• Static data — no API call, instant response; call once and reuse the IDs

All resource data is also reachable via tools. Use usda_search_foods to discover FDC IDs before reading food resources.

Built on @cyanheads/mcp-ts-core:

• Declarative tool and resource definitions — single file per primitive, framework handles registration and validation
• Unified error handling — handlers throw, framework catches, classifies, and formats
• Pluggable auth: none, jwt, oauth
• Swappable storage backends: in-memory, filesystem, Supabase, Cloudflare KV/R2/D1
• Structured logging with optional OpenTelemetry tracing
• STDIO and Streamable HTTP transports

USDA FDC-specific:

• Type-safe client for the USDA FoodData Central REST API (api.nal.usda.gov/fdc/v1)
• Normalization layer that handles inconsistent API response shapes across search, single-food, and batch endpoints
• Batch API endpoint (/foods) with per-food partial failure reporting
• HTML response detection (rate-limit proxy returns 200 HTML) with service-unavailable error surfacing
• Static nutrient reference dictionary — ~150 nutrients with FDC IDs, SR numbers, units, and categories; no API call required

Agent-friendly output:

• Provenance preserved — FDC data source (dataType) on every food result so callers know whether they're reading curated research data or label-derived branded values
• Partial failure reporting — batch tools (usda_get_foods, usda_compare_foods) return successes alongside structured failed[] / missingData[] entries rather than failing the whole request
• Cross-reference hints — FDC IDs and tool names in descriptions so agents know exactly which call to make next (search → get → compare workflow)

Add the following to your MCP client configuration file. See data.gov API key signup to generate a free API key.

{
  "mcpServers": {
    "usda": {
      "type": "stdio",
      "command": "bunx",
      "args": ["usda-mcp-server@latest"],
      "env": {
        "MCP_TRANSPORT_TYPE": "stdio",
        "MCP_LOG_LEVEL": "info",
        "USDA_FDC_API_KEY": "your-api-key"
      }
    }
  }
}

Or with npx (no Bun required):

{
  "mcpServers": {
    "usda": {
      "type": "stdio",
      "command": "npx",
      "args": ["-y", "usda-mcp-server@latest"],
      "env": {
        "MCP_TRANSPORT_TYPE": "stdio",
        "MCP_LOG_LEVEL": "info",
        "USDA_FDC_API_KEY": "your-api-key"
      }
    }
  }
}

Or with Docker:

{
  "mcpServers": {
    "usda": {
      "type": "stdio",
      "command": "docker",
      "args": [
        "run", "-i", "--rm",
        "-e", "MCP_TRANSPORT_TYPE=stdio",
        "-e", "USDA_FDC_API_KEY=your-api-key",
        "ghcr.io/cyanheads/usda-mcp-server:latest"
      ]
    }
  }
}

For Streamable HTTP, set the transport and start the server:

MCP_TRANSPORT_TYPE=http MCP_HTTP_PORT=3010 USDA_FDC_API_KEY=your-api-key bun run start:http
# Server listens at http://localhost:3010/mcp

• Bun v1.3.2 or higher (or Node.js v24+).
• A free USDA FDC API key — register at api.data.gov/signup. The key is required; without it the server will not start.

1. Clone the repository:

git clone https://github.com/cyanheads/usda-mcp-server.git

2. Navigate into the directory:

cd usda-mcp-server

3. Install dependencies:

bun install

4. Configure environment:

cp .env.example .env
# edit .env and set USDA_FDC_API_KEY

See .env.example for the full list of optional overrides.

• Build and run:

  # One-time build
  bun run rebuild
  
  # Run the built server
  bun run start:stdio
  # or
  bun run start:http

• Run checks and tests:

  bun run devcheck   # Lint, format, typecheck, security
  bun run test       # Vitest test suite
  bun run lint:mcp   # Validate MCP definitions against spec

docker build -t usda-mcp-server .
docker run --rm -e USDA_FDC_API_KEY=your-key -p 3010:3010 usda-mcp-server

The Dockerfile defaults to HTTP transport, stateless session mode, and logs to /var/log/usda-mcp-server. OpenTelemetry peer dependencies are installed by default — build with --build-arg OTEL_ENABLED=false to omit them.

See CLAUDE.md for development guidelines and architectural rules. The short version:

• Handlers throw, framework catches — no try/catch in tool logic
• Use ctx.log for request-scoped logging, ctx.state for tenant-scoped storage
• Register new tools and resources via the barrels in src/mcp-server/*/definitions/index.ts
• Wrap external API calls: validate raw → normalize to domain type → return output schema; never fabricate missing fields

Issues and pull requests are welcome. Run checks and tests before submitting:

bun run devcheck
bun run test

Apache-2.0 — see LICENSE for details.