Odel
w0 mcp server

w0 mcp server

@modern-ai-incJavaScriptUpdated 1w ago

Look up a brand's AI Recommendation Rate: how often it's the #1 AI pick.

View on GitHubSign in to connect
Server endpointStreamable HTTPNo authProbed
https://w0-mcp-server.modernai.workers.dev

This is the third-party server itself — Odel doesn't run it. Hitting this URL directly talks straight to the upstream server with no auth or proxying. Connect through Odel to front it with managed auth.

w0-mcp-server

A Model Context Protocol server for Modern AI's Brand Recommendation Rate lookup -- ask any MCP-compatible AI assistant (Claude, ChatGPT, etc.) how often a brand gets recommended by AI search and chat surfaces.

Live endpoint: https://w0-mcp-server.modernai.workers.dev Health check: GET /health -> ok Transport: MCP Streamable HTTP (single POST endpoint, JSON-RPC 2.0), hand-implemented, no SDK/build step.

What it exposes

One tool: lookup_brand_recommendation_rate(brand: string).

Given a brand name, returns:

  • Recommendation Rate -- the percentage of buyer-intent questions where the brand is the AI's #1 pick (not just mentioned).
  • Recommendation Inclusion Rate -- the percentage of buyer-intent questions where the brand appears anywhere in the answer.

This is a free, single-brand, rate-limited lookup backed by Modern AI's Discovery product. Brands that haven't been measured yet return an honest "not measured" response rather than a fabricated number.

Usage

Point any MCP-compatible client at https://w0-mcp-server.modernai.workers.dev using the Streamable HTTP transport. Example tools/call request body:

{
  "jsonrpc": "2.0",
  "id": 1,
  "method": "tools/call",
  "params": {
    "name": "lookup_brand_recommendation_rate",
    "arguments": { "brand": "Brumate" }
  }
}

Rate limits

Free lookups are capped per caller under Modern AI's published anti-scrape policy. A 429 response indicates the ceiling was reached for that session; Modern AI offers a commercial access tier with a higher ceiling.

Source

src/index.js is the complete Worker. It is a stateless pass-through: every call forwards to Modern AI's already-deployed, already rate-limited brand-lookup service, so no separate rate limit or attack surface is introduced by this server.

deploy.py deploys the Worker to Cloudflare via the direct REST API. It requires CLOUDFLARE_API_TOKEN, CLOUDFLARE_ACCOUNT_ID, and a shared W0_MCP_RELAY_SECRET set in the environment, and assumes the upstream gate Worker (w0-brand-gate) already exists in the same Cloudflare account with a matching secret.

More about Modern AI

Modern Discovery measures and improves how brands show up in AI search and chat answers. This server is one of several public integration points into that data.

Issues / notifications

This repository is monitored by Modern AI. Open an issue for bugs or questions.