# Fodda Brand Intelligence [Health: Active]

**Category:** 🎯 Marketing  
**Repository:** https://github.com/piers-fawkes/fodda-mcp  
**GitHub Stars:** 1  
**Views:** 0  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/fodda-brand-intelligence

## Description
Brand health & trend footprint across PSFK expert graphs with citable sources, not web summaries.

## Tools
Capabilities this server exposes over MCP:

- **list_graphs** — Discover available knowledge graphs, metadata, and routing instructions.
- **search_graph** — Hybrid keyword + semantic search across curated graphs for trend clusters and evidence.
- **search_statistics** — Search for specific quantitative metrics, indicators, and numeric data points.
- **search_insights** — Search expert quotes, qualitative signals, and professional interpretations.
- **get_label_values** — Discover valid values for a node label or category.
- **get_neighbors** — Traverse from seed nodes to discover related concept nodes and links.
- **get_node** — Retrieve detailed metadata, properties, and attributes for a single node by ID.
- **get_evidence** — Retrieve source signals, articles, citations, and provenance for a trend or node.
- **discover_adjacent_trends** — Find semantically similar trends to a given trend node.
- **get_domain_intelligence** — Search all PSFK curated domain graphs (retail, beauty, sports, fashion, consumer electronics, F&B) in parallel.
- **get_expert_intelligence** — Query specialist industry graphs built by leading strategists and experts.
- **get_report_intelligence** — Search institutional report insights from DHL, PwC, Delta, and other partners.
- **get_earnings_intelligence** — Query management commentary, business guidance, and Q&A from company earnings calls.
- **get_earnings_divergence** — Detect deflections, gaps, and divergence between analyst concerns and executive responses in earnings calls.
- **brand_tracker** — Compile a comprehensive Brand Intelligence Profile across all knowledge graphs.
- **brainstorm_topic** — Graph-powered brainstorm map discovering unexpected connections and adjacent territories.
- **deep_research_topic** — Launch an autonomous research session combining graphs with live web search.
- **check_research_status** — Check progress or retrieve the final narrative report of a deep research job.
- **get_supplemental_context** — Fetch macro context from up to 10 institutional data sources in a single query.
- **check_supplemental_status** — Retrieve output from a supplemental context job.
- **generate_comms_plan** — 2026 Strategy-to-Activation Comms & Retail Media Plan with T1-T5 tasks, 3-layer shopper audience cards, digital shelf intelligence, and activation manifest.
- **list_analysts** — List available Synthetic Analyst expert personas.
- **consult_analyst** — Engage a Synthetic Analyst persona to synthesize answers with specialized voice.
- **generate_visual** — Generate branded SVG data visualizations (Cultural Shifts, Competitive Compass, Innovation Pathway, etc.).

## Claude Desktop Quick Installation
Install path detected from listing signals. Uses `npx` (confidence: high):

```json
"mcpServers": {
  "fodda-brand-intelligence": {
    "command": "npx",
    "args": ["-y","fodda-mcp"],
    "env": {
      "PORT": "",
      "FODDA_API_URL": "",
      "FODDA_MCP_SECRET": "",
      "NODE_ENV": ""
    }
  }
}
```

**Requires environment variables:** `PORT`, `FODDA_API_URL`, `FODDA_MCP_SECRET`, `NODE_ENV` — the values above are empty placeholders; fill in real credentials before running (see the repository for what each one is for).

## Documentation

## What Fodda Brand Intelligence does

The Fodda Brand Intelligence MCP server gives AI agents access to curated knowledge graphs covering retail, beauty, sports, fashion, consumer electronics, and food and beverage. It combines trend clusters, expert commentary, quantitative indicators, institutional reports, company earnings material, and source provenance in one MCP interface.

Core discovery tools include `list_graphs`, `search_graph`, `search_statistics`, and `search_insights`. Agents can then inspect nodes, follow related concepts, find adjacent trends, and retrieve evidence such as articles, citations, and source signals.

For broader assignments, `brand_tracker` compiles a brand profile across knowledge graphs, while `brainstorm_topic` maps unexpected connections. `deep_research_topic` starts a research job that combines graph data with live web search; `check_research_status` returns progress or the completed narrative report.

## How it works

The MCP interface exposes tools through the hosted service at `https://mcp.fodda.ai/claude` or through a personal tokenized connection URL generated in the Fodda connections dashboard. Claude can authorize the hosted connector with Clerk OAuth. Header-capable clients can instead send a Fodda key as `Authorization: Bearer ...` or with `X-API-Key`.

Requests can begin with graph discovery and routing information, then narrow into keyword or semantic search. Deterministic operations such as node retrieval, neighbor traversal, label-value discovery, and evidence lookup can support a more inspectable workflow, while synthesis-oriented tools return interpreted intelligence. Earnings analysis includes management commentary, guidance, analyst questions, and possible divergence between analyst concerns and executive responses.

Additional tools retrieve macro context from up to 10 institutional data sources, consult named Synthetic Analyst personas, create communications and retail media plans, and generate branded SVG visualizations. Supplemental and deep research jobs have separate status tools for retrieving asynchronous results.

## Setup and configuration

For hosted Claude access, add the custom connector URL in Claude and complete OAuth authorization. Claude Code can register the same endpoint with HTTP transport, either through OAuth or an API-key bearer header. The repository points to Fodda’s connection guides for Claude, ChatGPT, OpenAI Responses API, Cursor, Gemini, and custom agents.

To run the repository locally, install dependencies, build the project, and start the service with npm. The documented configuration variables are `PORT`, `FODDA_API_URL`, `FODDA_MCP_SECRET`, and `NODE_ENV`; the defaults are port `8080`, the Fodda API URL, and production mode, while the HMAC secret has no listed default. Docker, Cloud Run, Kubernetes, and Terraform deployment paths are also documented.

## Limitations and notes

The Fodda Brand Intelligence MCP server depends on Fodda’s hosted account and authentication model when used remotely. Raw API keys are not accepted in connection URLs; use a tokenized connection URL or an HTTP authorization header instead.

The primary datasets are curated PSFK and partner sources, not an unrestricted web index. Although deep research can add live web search, individual graph searches are bounded by the available graphs and their indexed evidence. Partner-specific tools can also appear dynamically when additional account skills are enabled. The project is marked Proprietary, so confirm licensing terms before redistributing or modifying a deployment.

_Full upstream README: https://allmcps.com/mcp/fodda-brand-intelligence/readme_