agent-factory

You can automate an AI software development team by deploying 12 specialized agents that collaborate via GitHub Issues and Projects. This setup handles product ownership, architecture, development, testing, security, and infrastructure autonomously.

Autonomous project execution via GitHub Issues uses a state machine driven by labels and project boards to orchestrate multiple AI agents. It enables a full development team to build software from product vision to deployment without manual task delegation.

Set up an AI development team by running the provided setup script in your GitHub repository. This deploys the agent ecosystem, integrating with native GitHub tooling for workflow automation and state management to kickstart your project.

Yes, agent-factory integrates with GitHub Actions and native GitHub tooling for workflow automation. It includes a dedicated CI/CD agent to manage continuous integration and deployment pipelines alongside the other specialized agents.

Yes, you can use multi-agent AI for project management and code generation by deploying specialized PO, PM, and Developer agents. They collaborate through GitHub Issues to define product vision, manage architecture, and generate code autonomously.

Using AI agents for autonomous development requires full dependency on GitHub Issues and Projects for state management. The 12 specialized agents operate within this ecosystem, meaning workflow automation is constrained by GitHub's native tooling limits.