About this project

Mastra is a TypeScript framework for building AI-powered applications and agents, positioned as a way to move from prototypes to production-ready systems. It can be embedded in frontend or backend stacks such as React, Next.js and Node, or deployed as a standalone server. Core capabilities described in the README: - Model routing: a single standard interface for connecting to 40+ model providers, including OpenAI, Anthropic and Gemini. - Agents: autonomous components that use LLMs and tools, reason about goals, choose tools and iterate until a final answer or stopping condition. - Workflows: a graph-based engine for explicit control over multi-step processes, with control-flow syntax such as .then(), .branch() and .parallel(). - Human-in-the-loop: suspend an agent or workflow and wait for user input or approval, using storage to persist execution state so runs can pause and resume. - Context management: conversation history, retrieval from APIs, databases and files for RAG, plus an Observational Memory feature for more coherent agent behavior. - Integrations: bundle agents and workflows into React, Next.js or Node apps, or expose them as standalone endpoints; UI integration options include Vercel's AI SDK UI and CopilotKit. - MCP servers: author Model Context Protocol servers that expose agents, tools and structured resources to any MCP-compatible system. - Production essentials: built-in evals and observability for measuring and refining agents over time. Getting started is via `npm create mastra@latest`, with a CLI or manual installation path documented. A dev server exposes Mastra Studio at localhost:4111 for building, testing and managing agents, workflows and tools. Templates, a course and YouTube videos are referenced for newcomers. The project is dual-licensed: the core framework and most of the codebase are Apache-2.0, while code under directories named `ee/` is source-available under the Mastra Enterprise License, requiring a valid enterprise license for production use but free for development and testing. Security findings are requested to be reported to security@mastra.ai. Community support is available through an open Discord server.