Convex MCP Server

The Convex MCP Server is a utility server that integrates the Convex reactive database platform with AI assistants through the Model Context Protocol (MCP). This integration enables AI models to interact with Convex deployments, facilitating tasks such as data retrieval, function execution, and environment management. Key features include: - Deployment Introspection: Allows AI assistants to query available deployments and retrieve deployment selectors for subsequent operations. - Table Management: Enables listing of all tables within a deployment, including their declared and inferred schemas, and supports pagination through table documents. - Function Execution: Provides metadata about deployed functions and allows AI models to execute these functions with specified arguments. - Sandboxed Query Execution: Supports writing and executing read-only JavaScript queries against the deployment's data in a secure, sandboxed environment. - Environment Variable Manipulation: Facilitates listing, retrieving, setting, and removing environment variables within a deployment. Designed for local operation, the server requires Node.js and the Convex CLI for execution. It is actively maintained as an open-source project.