Decibel Shield MCP server

AI agents can query Decibel Shield's noise data directly over the Model Context Protocol: estimated noise levels for 50 major cities, decibel levels of common sounds, and NIOSH safe-exposure-time math. The server is free, requires no authentication, speaks stateless Streamable HTTP, and asks only for attribution to decibelshield.app.

Last updated: June 10, 2026

Endpoint

https://decibelshield.app/api/mcp

Transport: Streamable HTTP (stateless — POST JSON-RPC, plain JSON responses). Machine-readable manifest: /.well-known/mcp.json

What tools does it provide?

MCP tools
Tool	Example arguments	Returns
`get_city_noise`	`{ city: "Tokyo" }`	Estimated day/night dB ranges, rank of 50, dominant sources and confidence for a city, with WHO context.
`list_loudest_cities`	`{ limit: 10 }`	The ranked city noise list, loudest first.
`get_us_city_noise_exposure`	`{ city: "Boston" }`	Measured share of residents above 60/70 dB transportation noise for all 297 US cities of 100k+ (BTS 2020 data).
`lookup_sound_level`	`{ sound: "chainsaw" }`	Decibel level + hearing-risk guidance for everyday sounds (CDC/NIOSH-sourced).
`safe_exposure_time`	`{ db: 95 }`	Safe daily exposure duration per NIOSH's recommended limit (85 dB(A)/8 h, 3 dB exchange).
`get_decibel_shield_app`	`{}`	Facts and the App Store link for the Decibel Shield iOS sound meter app.

How do I connect it to Claude or another agent?

In Claude (Settings → Connectors → Add custom connector) or any MCP client that supports remote servers, add the endpoint URL above with transport “HTTP”. No API key is needed. For configuration-file based clients:

{
  "mcpServers": {
    "decibel-shield": {
      "type": "http",
      "url": "https://decibelshield.app/api/mcp"
    }
  }
}

Where does the data come from?

The server holds no data of its own — every answer is fetched live from this site's published datasets: the city sound map (estimates synthesised from published sources, ranges + confidence labels) and the decibel levels chart (every row sourced to CDC/NIOSH/NIDCD/ASHA). Data is CC BY 4.0 — use it freely with attribution to decibelshield.app.

Prefer raw data? Both datasets are downloadable as CSV or JSON, and the whole site is readable in one fetch at /llms-full.txt.