Compozy is an agent orchestration framework designed for building and deploying multi-agent AI systems. It targets developers and enterprises seeking a scalable, reliable, and cost-efficient platform for complex AI workflows, simplifying distributed task execution, debugging, and monitoring.

How It Works

Compozy utilizes a declarative approach, allowing users to define AI workflows using YAML. It leverages Temporal for robust backend orchestration, ensuring persistence, monitoring, and security. The framework supports a variety of task types, including parallel, sequential, and conditional execution, and is extensible with custom tools written in TypeScript/JavaScript. Its core is built with Go for high performance.

Quick Start & Requirements

Installation is available via Homebrew (brew install compozy) or package managers like NPM, Yarn, or PNPM (npm install -g @compozy/cli). A quick start involves initializing a project (compozy init my-ai-app) and starting the development server (compozy dev). Further details can be found in the Quick Start Guide.

Highlighted Details

• Supports 7+ LLM providers (OpenAI, Anthropic, Google, local models).
• Features 8 powerful task types for advanced orchestration.
• Developer-focused CLI with hot-reloading.
• Built with Go for high performance and Temporal for enterprise-grade reliability.

Maintenance & Community

Compozy welcomes community contributions via GitHub Issues for bug reports and feature requests. A contributing guide is available.

Licensing & Compatibility

The project is licensed under the Business Source License 1.1 (BUSL-1.1), which may have restrictions on commercial use or redistribution.

Limitations & Caveats

The project is currently in an alpha release, indicating potential instability, bugs, and significant changes. Users should exercise caution.