Summary

Hatchify is a multi-agent workflow platform addressing complex AI agent collaboration. It targets engineers and researchers building sophisticated AI systems, offering dynamic graph execution and orchestration via JSON configuration, simplifying the development of interconnected AI agents.

How It Works

Built on FastAPI and AWS Strands SDK, Hatchify employs a dynamic graph execution engine. It enables workflow creation and execution through JSON configurations, featuring an intelligent routing system (Rules, JSONLogic, Router Agent) and native Model Context Protocol (MCP) integration for tool extensibility. An event-driven architecture with SSE facilitates real-time execution tracking.

Quick Start & Requirements

• Install: Clone the repository, cd hatchify, then uv sync.
• Prerequisites: Python 3.13+, Node.js 18+ (for Web Builder). SQLite is the only currently supported database.
• Configuration: Copy and edit resources/mcp.toml and resources/models.toml with API keys and server details.
• Launch: Use uvicorn hatchify.launch.launch:app --reload or python main.py. Docker deployment is also supported.
• Links: Official Website, GitHub. Documentation is pending.

Highlighted Details

• Vibe Graph: Natural language-driven workflow generation using LLMs to translate requirements into executable GraphSpec definitions, supporting conversational iteration.
• Flexible Graph Building: Supports declarative graphs with Agent Nodes (LLM-based) and Function Nodes (deterministic operations), allowing dynamic model selection and tool integration (MCP, custom local tools).
• MCP Protocol Integration: Native support for Model Context Protocol enables dynamic loading of external tool servers (e.g., filesystem, git).
• Web Builder (In Development): AI-powered conversational web application generation, from concept to deployment, using React, TypeScript, and Tailwind CSS.
• Unified Configuration: Centralized management of LLM models and MCP tools via declarative TOML and YAML files, supporting multiple providers and transport protocols.

Maintenance & Community

No specific details regarding maintainers, community channels (e.g., Discord, Slack), or a public roadmap are provided in the README.

Licensing & Compatibility

• License: MIT License.
• Compatibility: The MIT license permits commercial use and integration into closed-source projects.

Limitations & Caveats

The Web Builder feature is currently in development. Database support is limited to SQLite, with PostgreSQL and MySQL planned for future releases. Official documentation is not yet available. The server.base_url configuration requires explicit modification for production deployments.