This readme documents key requirements and patterns for the Postman API MCP server handlers.
Visit modelcontextprotocol.io/tutorials/building-mcp-with-llms for a comprehensive guide covering:
- Documentation preparation
- Server development workflow
- Implementation best practices
- Testing and deployment steps
- Prompt IDs must be used in snake_case format (e.g., 'create_collection', not 'Create Collection')
- Input validation is performed using TypeScript type guards
- Each prompt has a defined input schema and generates appropriate messages
- Handles direct resource URIs in the format:
postman://{resource-type}
- Resource types are fixed strings without parameters (e.g., 'workspaces', 'collections')
- Direct resources provide top-level API data (e.g., list of all workspaces)
- Handles parameterized URIs in the format:
postman://{resource-type}/{id}/[sub-resource]
- Supports nested resources (e.g., workspace collections, API versions)
- Template parameters are validated before making API requests
postman://workspaces # List all workspaces
postman://collections # List all collections
postman://environments # List all environments
postman://workspaces/{id}/collections # Collections in a workspace
postman://apis/{id}/versions # Versions of an API
postman://collections/{id}/requests # Requests in a collection
For implementation examples and detailed guidance, refer to the official tutorial.