MarcAntoineIseg/eliott-mcp-server

MCP Server for Eliott (AI Marketing)

This backend exposes two MCP-compliant endpoints for use by AI agents (OpenAI tool_call):

• POST /search — Search Google Ads campaigns for a user
• POST /fetch — Fetch a campaign's details by ID

Tech stack

• Node.js + Express
• Supabase (OAuth2 token storage)
• Google Ads API (OAuth2)
• axios
• supabase-js

Project structure

/mcp-server
├── package.json
├── index.js
├── routes/
│   └── search.js
│   └── fetch.js
├── services/
│   └── googleAds.js
│   └── supabaseClient.js
├── utils/
│   └── formatMCPResponse.js
├── .env.example

Setup

1. cp .env.example .env and fill in your values
2. npm install
3. npm start

Environment variables

See .env.example for required variables.

Deployment

Ready for Railway deployment. The start script is set up for production.

---

Endpoints:

POST /search

• Body: { query: string, uid: string }
• Returns: Array of MCP results ([{ id, title, text, url }])

POST /fetch

• Body: { id: string }
• Returns: MCP document ({ id, title, text, url, metadata })

---

For any issues, see the code comments for extension points and details.