Square

Exposes Square's complete API service catalog through the get_service_info MCP tool, enabling AI assistants to programmatically discover available services (payments, customers, inventory, etc.) and their methods without manual documentation lookup. The server maintains an in-memory registry of Square API services and returns structured metadata about available operations, parameters, and return types through the standardized MCP tool interface.

The get_type_info MCP tool resolves detailed parameter requirements and data structure definitions for Square API methods, translating Square's OpenAPI-derived type system into structured schemas that AI assistants can use for request validation and construction. This tool returns comprehensive type metadata including required fields, field types, constraints, and nested object structures, enabling AI systems to construct valid API payloads without trial-and-error.

The make_api_request MCP tool executes authenticated HTTP requests to Square's Connect API, handling credential injection (API key from environment), request serialization, and response parsing through a single MCP interface. The server manages API authentication state via environment variables (SQUARE_ACCESS_TOKEN) and abstracts away HTTP client details, allowing AI assistants to invoke Square operations by specifying service, method, and parameters without managing authentication or network concerns.

Provides a command-line interface (CLI) for installing, starting, and configuring the Square MCP server for use with AI assistants (Claude, Goose). The CLI handles server initialization, environment variable setup, and generates integration URLs for connecting AI assistants to the running server. The server implements the MCP protocol specification, managing tool registration, request routing, and response serialization for the three core tools (get_service_info, get_type_info, make_api_request).

Exposes a comprehensive catalog of Square's business API services organized across 20+ domains including payments, inventory, customers, bookings, labor, and hardware integration. The server maintains structured metadata for each service category (Catalog & Inventory, Customers & Orders, Payments & Financial, Business Management, etc.), enabling AI assistants to discover and operate across Square's entire business platform without domain-specific knowledge.

Provides pre-built integration paths for Claude (via Anthropic API) and Goose AI assistants, with automatic configuration generation and URL-based setup. The server detects the target AI assistant and generates appropriate integration URLs or configuration snippets, abstracting away MCP protocol details and enabling one-command setup for connecting AI assistants to Square's API ecosystem.

Manages Square API authentication and server configuration through environment variables (SQUARE_ACCESS_TOKEN, etc.), providing a secure, externalized credential store that isolates secrets from code and configuration files. The configuration system reads environment variables at server startup and injects credentials into API requests, enabling secure credential management without exposing keys to AI assistants or storing them in version control.

Implements the Model Context Protocol (MCP) specification using JSON-RPC 2.0 messaging, with automatic tool registration and request routing for the three core Square tools. The server handles MCP protocol details including request/response serialization, error handling, and tool discovery, abstracting away protocol complexity from AI assistants and enabling them to invoke Square operations through a standardized interface.