Paca offers a free, open-source, and self-hosted alternative to popular project management tools like Jira and Trello, specifically designed for Scrum teams. Its core innovation is enabling AI agents to function as first-class teammates alongside humans on the same boards and sprints, fostering genuine collaboration rather than peripheral automation. This AI-native approach aims to democratize advanced project management capabilities for all teams.

How It Works

Paca's central philosophy is integrating AI agents directly into the Scrum process. These agents participate in sprint planning, pick up tasks, write BDD specs, and contribute to system design documents, mirroring human team members. The platform is built on a lightweight core that is fully customizable via configuration files for workflows, statuses, and fields, and extendable through a plugin system. Plugins can be developed in languages with WASM targets for the backend or as standard modules for the frontend, running in a sandboxed environment.

Quick Start & Requirements

The recommended installation method is an interactive script that runs on any Linux server with Docker installed: curl -fsSL https://github.com/Paca-AI/paca/releases/latest/download/install.sh | bash. Alternatively, a Docker Compose setup is available. Key prerequisites include Docker. The architecture also involves PostgreSQL for persistence and Valkey for caching and event streams.

Highlighted Details

• AI agents act as first-class Scrum teammates on a unified, real-time Scrumban board.
• Fully customizable through project-level configuration files and a plugin system (WASM/JS).
• Self-hosted architecture ensures data ownership and privacy, offered under the Apache 2.0 license.
• Includes an MCP (Model Context Protocol) server to connect any compatible AI agent, such as Claude Desktop.
• Features a Claude Code skill for managing Paca workspaces directly from an editor via slash commands.
• Provides an activity diff and revert feature for granular change tracking and rollbacks.
• Facilitates BDD collaboration with AI agents for writing Gherkin scenarios and SDDs.

Maintenance & Community

The provided README does not detail specific contributors, sponsorships, or community channels like Discord or Slack. A roadmap is mentioned, but no direct link is supplied.

Licensing & Compatibility

Paca is distributed under the permissive Apache License 2.0. This license allows for commercial use and integration into closed-source projects without significant restrictions, making it highly compatible for various business needs.

Limitations & Caveats

The README focuses on features and setup, not explicit limitations. As a self-hosted solution, it requires users to manage Docker and associated infrastructure. AI agent integration depends on external MCP-compatible clients or specific integrations like Claude Code. Developing custom plugins requires familiarity with WASM or JavaScript module development.