Fodda MCP Server

Expert-curated knowledge graphs for AI agents — 100+ datasets across retail, beauty, sports, culture and more via the Model Context Protocol.

Your AI produces generic output because it has generic context. Fodda gives it expert-curated knowledge and real institutional evidence — so it reasons better and explains its work.

Unlike flat text or generic retrieval, Fodda organizes insight as interconnected entities, relationships, evidence, and metadata — giving AI systems both the qualitative context and quantitative evidence they need for trusted, explainable outputs.

Quick Start · Available Tools · Graph Showcase · How It Works · REST API · Best Practices

Quick Start

After receiving your API key (sk_live_...), connect to Fodda in your preferred tool:

Claude (Pro / Max / Team / Enterprise)

Connector URL: https://mcp.fodda.ai/mcp?api_key=YOUR_API_KEY&user_id=YOUR_EMAIL

Paste into: Claude → Settings → Connectors → Add custom connector. Leave OAuth fields blank.

Claude Code (CLI)

claude mcp add --transport sse fodda https://mcp.fodda.ai/sse \
  --header "Authorization: Bearer YOUR_API_KEY"

Claude Desktop

Add to claude_desktop_config.json:

{
  "mcpServers": {
    "fodda": {
      "url": "https://mcp.fodda.ai/sse",
      "headers": {
        "Authorization": "Bearer YOUR_API_KEY"
      }
    }
  }
}

Cursor / Windsurf / Any MCP Client

{
  "mcpServers": {
    "fodda": {
      "url": "https://mcp.fodda.ai/sse",
      "headers": {
        "Authorization": "Bearer YOUR_API_KEY"
      }
    }
  }
}

Notion (Business / Enterprise)

1. Custom Agent → Tools & Access → Add connection → Custom MCP server

2. URL: https://mcp.fodda.ai/mcp?api_key=YOUR_API_KEY&user_id=YOUR_EMAIL

3. Auth type: API Key → Key: api_key, Value: YOUR_API_KEY

Microsoft 365 Copilot

Use the Agents Toolkit in VS Code → Add an Action → Start with an MCP Server → URL: https://mcp.fodda.ai/sse

Google Gemini (via Vertex AI)

{
  "tools": [{
    "type": "mcp",
    "name": "fodda",
    "url": "https://mcp.fodda.ai/sse",
    "headers": {
      "Authorization": "Bearer YOUR_API_KEY",
      "X-Fodda-Mode": "deterministic"
    }
  }]
}

OpenAI Codex / ChatGPT (via REST API)

# List available graphs (public, no auth required)
curl https://api.fodda.ai/v1/graphs/catalog

# Search a graph
curl -X POST https://api.fodda.ai/v1/graphs/retail/search \
  -H "X-API-Key: YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"query": "checkout friction", "limit": 10, "use_semantic": true}'

No API key yet? Try Fodda instantly by adding one of these trial connectors in Claude:

Graph	Trial Connector URL
🛒 Retail	https://mcp.fodda.ai/mcp?api_key=sk_trial_retail&user_id=claude-user
💄 Beauty	https://mcp.fodda.ai/mcp?api_key=sk_trial_beauty&user_id=claude-user
⚽ Sports	https://mcp.fodda.ai/mcp?api_key=sk_trial_sports&user_id=claude-user

Available Tools

Once connected, your agent has access to these tools:

When using Fodda via MCP, your agent has access to 22 live institutional data sources spanning US economic indicators, health & science databases, global trade data, academic research, and consumer market signals. These are accessed via the unified get_supplemental_context tool — you don't need to call individual sources directly.

Discovery Endpoints

Graph Showcase

Fodda hosts 100+ expert-curated knowledge graphs. Here are some highlights:

See all graphs: curl https://api.fodda.ai/v1/graphs/catalog

How It Works

Your Agent (Claude, Cursor, Gemini, Copilot, etc.)
        ↓ calls MCP tools
Fodda MCP Server (mcp.fodda.ai)
        ↓ translates to REST calls
Fodda API (api.fodda.ai)
        ↓ queries
Neo4j Knowledge Graphs + Supplemental Sources

What happens during a typical prompt

When you ask your agent a research question, the MCP server orchestrates multiple tool calls behind the scenes:

1. list_graphs — Agent discovers which graphs are relevant to your query

2. search_graph — Fires a hybrid semantic + keyword search across one or more graphs

3. get_evidence — Retrieves source articles backing the top trends

4. search_statistics — Pulls curated metrics, expert quotes, and signals

5. get_supplemental_context — The unified supplemental tool fires relevant institutional data sources based on your query content, returning quantitative evidence across economic, health, academic, and market categories

6. get_adjacent_trends — Discovers related trends across graph boundaries

A single prompt typically chains 6–10 tool calls — this is why 1 MCP prompt ≈ 8 API calls.

Supplemental data

The unified supplemental tool intelligently routes your query to the most relevant institutional data sources based on topic and domain. Results are consolidated into five stable categories — demand signals, economic context, market data, research signals, and demographic context — and woven into the agent's response with full source attribution.

API Call Usage

Check balance with the get_my_account tool or at app.fodda.ai.

REST API Reference

Base URL: https://api.fodda.ai

All private requests require: X-API-Key: YOUR_API_KEY

Public Endpoints (no auth required)

• GET /v1/graphs/catalog — Public graph registry

• POST /v1/psfk/overview — LLM-synthesized industry overview

Core Graph Endpoints (API key required)

• POST /v1/graphs/:graph_id/search — Hybrid semantic + keyword search

• GET /v1/graphs/:graph_id/nodes/:node_id — Node metadata lookup

• POST /v1/graphs/:graph_id/neighbors — Graph traversal

• POST /v1/graphs/:graph_id/evidence — Source articles for a trend

• GET /v1/graphs/:graph_id/statistics — Curated metrics, quotes, signals

• GET /v1/graphs/:graph_id/adjacent — Semantically similar trends

• POST /v1/brand-intelligence/:brandName — Cross-graph brand analysis

Multi-Graph Search Endpoints (API key required)

• POST /v1/search/domain — Search all curated domain graphs

• POST /v1/search/expert — Search all expert specialist graphs

• POST /v1/search/report — Search all industry report graphs

• POST /v1/search/filtered — Semantic search with metadata filters

Research Endpoints (API key required)

• POST /v1/research/deep-dive — Autonomous deep-dive research

• POST /v1/research/chat — Multi-turn research with session memory

Copilot Adapter Endpoints (API key required)

• POST /copilot/search_insights — Flattened narrative insights

• POST /copilot/get_evidence — Trend-level supporting articles

• POST /copilot/get_statistics — Quantitative statistics search

Full API docs: app.fodda.ai/knowledge/api-docs

Authentication

Pass your Fodda API key as a Bearer token:

Authorization: Bearer sk_live_...

Troubleshooting

Best Practices

1. Use thematic, conversational queries: Full sentences work best — e.g., "How are retailers removing friction from the buying journey?" outperforms short keywords like "checkout friction"

2. Search, then deep-dive: First search_graph, then use get_evidence and search_statistics on specific trends

3. Cross-graph discovery: Use get_adjacent_trends to find connections between domains, or search_domain_graphs to search across all curated graphs at once

4. Validate with data: Use get_supplemental_context to add quantitative backing from institutional data sources

5. Use filtered search: When you know what you're looking for, use search_filtered with category, date, brand, or geography filters for precision results

MCP Registry

This server is published to the Official MCP Registry as io.github.piers-fawkes/fodda.

# Verify listing
curl "https://registry.modelcontextprotocol.io/v0.1/servers?search=io.github.piers-fawkes/fodda"

Links

• Website: fodda.ai

• App: app.fodda.ai

• API Docs: app.fodda.ai/knowledge/api-docs

• QuickStart Guide: Fodda_Quickstart.md

• npm: fodda-mcp

• Support: piers.fawkes@psfk.com

License

Proprietary — fodda.ai