Revinho

What Revinho MCP server does

Revinho MCP server connects an AI agent to business intelligence collected in a Revinho workspace. The hosted service combines Google Analytics 4 data, Stripe data, and optionally Google Search Console data. Its read-only interface is intended for questions about performance changes, acquisition, conversion, revenue, page effectiveness, and areas that deserve attention.

The server exposes six focused tools:

• get_overview returns headline performance, comparisons between periods, trends, acquisition context, freshness information, and deterministic signals.
• get_growth examines channels, sources, campaigns, countries, organic queries, and acquisition segments.
• get_funnels reports saved GA4 journeys and performance at each funnel step.
• get_pages covers page traffic, key events, conversion, search demand, classifications, and related impact.
• get_revenue provides Stripe-based revenue, MRR, subscriptions, customers, refunds, plans, and currency metadata.
• get_opportunities ranks deterministic opportunities and includes supporting evidence and attribution provenance.

Clients that support MCP prompts may also use investigate_business, which leads the agent through a bounded investigation and stops when the available evidence answers the question.

How it works

Revinho MCP server is operated as a hosted production service; the GitHub repository is used for discovery, connection information, and support rather than deploying the production server. The MCP endpoint is https://www.revinho.com/api/mcp, and communication uses Streamable HTTP. Requests are authorized with a bearer token scoped to a Revinho workspace.

All tools are read-only, non-destructive, idempotent, and closed-world. Revenue comes from Stripe as the source of truth. When revenue is linked to acquisition or behavior, the response preserves whether the attribution is verified, modeled, or associated. Relationships that do not establish causation are described as coinciding or associated. Provider freshness fields indicate data age, and missing or partial values are not silently converted to zero.

Setup and configuration

Create a connection by signing in to Revinho, opening Setup your agent, and creating a named workspace token. Tokens start with rv_live_ and are displayed only once. Configure the hosted endpoint in an MCP-compatible client with the token in an Authorization header:

{
  "mcpServers": {
    "revinho": {
      "type": "streamable-http",
      "url": "https://www.revinho.com/api/mcp",
      "headers": {
        "Authorization": "Bearer rv_live_REPLACE_WITH_YOUR_TOKEN"
      }
    }
  }
}

Client configuration syntax varies. Use the client’s secret or environment-variable facility where available, and do not commit the token to source control. Revinho MCP server does not require a locally deployed process according to the provided setup instructions.

Limitations and notes

Revinho MCP server only provides read access, so it cannot modify analytics, billing, search-console, or workspace data. Its conclusions are limited to information available in the connected workspace and the evidence returned by the providers. Opportunities are generated by deterministic rules and stored evidence rather than unspecified recommendations.

Use separate tokens for agents or environments, keep them in a secret manager or private client configuration, and revoke a token if exposure is suspected. Tokens are workspace-scoped, read-only, stored as cryptographic hashes, and revocable. The repository’s documentation and metadata are MIT-licensed, but that license does not grant the hosted service, product code, name, or brand assets.