@verityscore/mcp-server

MCP (Model Context Protocol) server for Verity Score — the Shopify-first GEO audit platform.

5 Tools

Tool	Description
`get_geo_score`	GEO readiness score for a Shopify store (8 dimensions, top findings, report link)
`check_ai_readiness`	Real-time check of 5 AI discovery files (robots.txt, llms.txt, ai.txt, agent-card.json, sitemap.xml)
`get_recommendations`	Top 3 priority fixes with estimated revenue impact
`explain_topic`	Expert guides on 19 GEO & AI commerce topics
`get_vertical_info`	Checklists & benchmarks for 15 e-commerce verticals

Quick Start

Use the hosted server (recommended)

Add to your Claude Desktop or MCP client config:

config.json

{
  "mcpServers": {
    "verity-score": {
      "url": "https://mcp.verityscore.io/mcp"
    }
  }
}

Run locally

Terminal

npx @verityscore/mcp-server

The server starts on http://localhost:3100/mcp.

Test with curl

bash

# List all tools
curl -X POST http://localhost:3100/mcp \
  -H "Content-Type: application/json" \
  -H "Accept: application/json, text/event-stream" \
  -d '{"jsonrpc":"2.0","method":"tools/list","id":1}'

# Check AI readiness of any site
curl -X POST http://localhost:3100/mcp \
  -H "Content-Type: application/json" \
  -H "Accept: application/json, text/event-stream" \
  -d '{"jsonrpc":"2.0","method":"tools/call","params":{"name":"check_ai_readiness","arguments":{"domain":"mybrand.com"}},"id":2}'

Architecture

Transport: Streamable HTTP (stateless)
Auth: None (rate-limited to 10 req/min/IP)
Protocol: MCP v2025-03-26
SDK: @modelcontextprotocol/sdk v1.x

Tools Detail

`get_geo_score`

Returns the GEO audit score for a Shopify store. If the store has been audited, returns scores across 8 dimensions + top 5 issues. If not yet audited, queues it for analysis within 72 hours.

Code

Input:  { "domain": "mybrand.com" }
Output: { "status": "audited", "scores": [...], "topIssues": [...], "reportUrl": "..." }

`check_ai_readiness`

Real-time check of 5 AI discovery files. No database needed — fetches files directly from the target domain.

Code

Input:  { "domain": "mybrand.com" }
Output: { "score": 3, "maxScore": 5, "checks": [...], "recommendation": "..." }

`get_recommendations`

Returns top 3 actionable fixes sorted by revenue impact.

Code

Input:  { "domain": "mybrand.com" }
Output: { "totalIssues": 42, "topRecommendations": [{ "priority": 1, "title": "...", "fix": "..." }] }

`explain_topic`

Expert knowledge base covering 19 GEO topics. Fuzzy-matches by slug, tag, or title.

Code

Input:  { "topic": "schema.org" }
Output: { "title": "Schema.org Product: Why and How", "keyPoints": [...], "readMore": {...} }

`get_vertical_info`

Checklists and benchmarks for 15 e-commerce verticals. Without argument, lists all verticals.

Code

Input:  { "vertical": "beauty" }
Output: { "checklist": { "requiredContent": [...], "expectedTrust": [...] }, "benchmarks": {...} }

License

MIT

Geo Audit README

@verityscore/mcp-server

5 Tools

Quick Start

Use the hosted server (recommended)

Run locally

Test with curl

Architecture

Tools Detail

`get_geo_score`

`check_ai_readiness`

`get_recommendations`

`explain_topic`

`get_vertical_info`

Links

License