MCP Server (beta)
We're excited to announce that the Stigg MCP Server is now available in Public Beta.
🆕 What’s new
You can now expose your Stigg environment directly to any AI agent or coding assistant that supports the Model Context Protocol (MCP) - including Claude Desktop, Cursor, ChatGPT, Windsurf, and more. Once connected, your AI assistant can interact with the entire Stigg API surface using natural language.
⭐️ Why it matters
Building and testing pricing models often requires switching between code, API documentation, and the Stigg UI. With the Stigg MCP Server, you can streamline your development and testing workflows directly from your IDE or AI chat interface without writing a single API call manually.
Your AI agent can now securely act on your behalf to:
- Manage your catalog: Create, update, and list plans, add-ons, and features.
- Handle customers & subscriptions: Provision customers, preview or create subscriptions, and grant promotional entitlements.
- Test usage & entitlements: Check if a customer has access to a specific feature, report metered usage, or submit raw metering events for testing.
⚒️ Get started
The Stigg MCP Server connects to your AI client using a standard HTTP transport or via our NPM package, authenticated securely via your Stigg environment API keys.
To connect your favorite AI assistant:
- Generate an API key in the Stigg app under Integrations > API keys (we highly recommend creating a dedicated, scoped key for your AI client).
- Configure your specific AI client (like Cursor, Claude, or ChatGPT) to point to the Stigg MCP Server.
- Start asking your agent to interact with Stigg! (e.g., "Check if customer X has access to the premium feature" or "Create a new test customer and subscribe them to the Pro plan").
For step-by-step configuration snippets for all major AI clients, check out the Stigg MCP Server documentation.