revenuecat-mcp

The analytics layer RevenueCat's official MCP is missing.

RevenueCat's official MCP server (launched July 2025) gives you 26 tools to build your subscription infrastructure from the terminal. It has zero tools to understand it.

This server fills that gap. It connects RevenueCat's Charts API to any AI agent — Claude, Cursor, Windsurf, and more — so instead of logging into the dashboard, you can ask:

"How is my MRR trending over the last 6 months?" "What's my current churn rate and is it healthy?" "Run a full subscription health check."

And get a real answer, backed by live data.

Why this matters now: RevenueCat's 2026 State of Subscription Apps found AI-powered apps earn 41% more per user but churn 30% faster. If you're building an AI app, analytics visibility isn't optional.

Tools

| Tool | What it does | |------|-------------| | get_overview | Snapshot of all key metrics: MRR, revenue, subscribers, trials, new customers | | get_chart | Time-series data for any of 21 available charts | | get_chart_options | Discover available resolutions, segments, and filters for a chart | | analyze_subscription_health | Full health report: MRR trend + churn + trial conversion with 🟢🟡🔴 indicators |

Available Charts (21 total)

mrr · revenue · actives · arr · churn · conversion_to_paying · customers_new · ltv_per_customer · ltv_per_paying_customer · mrr_movement · refund_rate · subscription_retention · subscription_status · trials · trials_movement · trials_new · actives_movement · actives_new · customers_active · trial_conversion_rate · cohort_explorer

Requirements

• Node.js 18+
• A RevenueCat V2 secret key with charts_metrics:charts:read and charts_metrics:overview:read permissions
• Your RevenueCat project ID (find it in Project Settings → API Keys)

Setup

Claude Desktop

Add to your claude_desktop_config.json:

{
  "mcpServers": {
    "revenuecat": {
      "command": "npx",
      "args": ["-y", "revenuecat-analytics-mcp"]
    }
  }
}

Cursor / Windsurf / Other MCP clients

{
  "mcpServers": {
    "revenuecat": {
      "command": "node",
      "args": ["/path/to/revenuecat-mcp/dist/index.js"]
    }
  }
}

Run locally

git clone https://github.com/gbrussich52/revenuecat-mcp
cd revenuecat-mcp
npm install
npm run build
node dist/index.js

Usage Examples

Once connected, prompt your AI agent naturally:

What's the current MRR for project proj058a6330?
→ Calls get_overview with your project ID and API key

Show me monthly revenue for the last year.
→ Calls get_chart(chart_name="revenue", start_date="2025-04-01", resolution="2")

Run a subscription health check.
→ Calls analyze_subscription_health — pulls MRR trend, churn, and trial conversion

What's my churn rate segmented by country?
→ Calls get_chart(chart_name="churn", segment="country")

Getting Your Credentials

1. Go to RevenueCat Dashboard → Project Settings → API Keys
2. Create a new V2 Secret Key
3. Enable permissions: charts_metrics:charts:read + charts_metrics:overview:read
4. Your Project ID is in the URL: app.revenuecat.com/projects/YOUR_PROJECT_ID/

Architecture

Your AI Agent (Claude / Cursor / etc.)
        ↓  MCP tool call
  revenuecat-mcp (this server)
        ↓  HTTPS
  RevenueCat Charts API V2
        ↓
  Your subscription data

The server acts as a translation layer: it takes natural-language-adjacent tool calls from your AI agent, maps them to the correct RevenueCat API endpoints, and returns clean markdown-formatted results.

Rate Limits

RevenueCat's Charts API has a 5 requests/minute rate limit on the charts domain. The analyze_subscription_health tool makes 4 parallel requests — be mindful if calling it frequently.

Built by

This server was built by Ainsley, an AI agent, as part of a take-home assignment for the RevenueCat Agentic AI Developer & Growth Advocate role. All code was generated autonomously using Claude Code.

Ainsley is an AI agent built on OpenClaw. This disclosure is required — and proudly made.

License

MIT