Introduction

OpenFlows is an autonomous AI software team that turns GitHub issues into reviewed, production-ready pull requests — running entirely inside your own self-hosted Coder environment. This section is where you start, whether you are just exploring or deciding whether it fits your team.

The reason it exists: writing code is no longer the bottleneck in software teams. Agents can produce code faster than any person can review it, so the first-pass reviewer has to be another agent and people have to move up the chain — to the architecture and product decisions only humans can own. OpenFlows automates the repetitive, low-judgment work in a single well-defined pipeline, from issue to merged pull request, so humans direct instead of babysit.

It is engineered as an enterprise tool — self-hosted, governed, auditable, and multi-tenant — and it works just as well for a solo developer who wants a personal AI team running their backlog. Whether you are one person or a company with many teams, humans stay at the architecture and product level while the agents coordinate the development.

The founding idea runs through everything you will read: Coder governs where agents run; OpenFlows governs how they coordinate. Coder provides the governed, disposable workspaces and the model gateway. OpenFlows provides the team, the planning gate, adversarial review, self-healing, and merges. The guardrails that shape the product are constant: plan before code, when in doubt don't approve, bounded recovery then human escalation, and no secrets in worker workspaces.

Where to go next

• Ready to run it? Jump to Quick Start.
• Understand the ideas? Read The Coder / OpenFlows Split.
• Evaluate the architecture? Start with Big Picture and Security by Design.