About

We make software and set it free.

Everything we build depends on work that came before it — other people’s ideas, tools, and generosity. We carry that forward: whatever we make is there to fuel whoever comes next.

The communal tradition that built most of the software we rely on is being displaced. Open source has become a supply chain. Community has become a marketing vertical. “Contribution” means something different to a VC than a maintainer. The culture has become extractive.

Free as in Bird is us continuing the old ways — building tools, sharing knowledge, and taking our obligation to the community that got us here seriously.

On Licensing

We believe the knowledge embedded in source code belongs to everyone. Our licenses protect that belief. What you build with this shared knowledge is yours.

Software is a communal craft. Every project depends on work that came before it — on other people’s ideas, discoveries, creations, and generosity. We choose licenses that keep that chain of knowledge open, so anyone who comes after us can learn from, use, and improve what we’ve built.

The moat in software is execution — infrastructure, platform, service, craft. It doesn’t need to be at the level of ideas. If someone builds something valuable using our work, that value belongs to them. But if someone improves the knowledge itself — fixes a bug, refines an algorithm, extends a pattern — that improvement belongs to everyone.

We match the license to the type of work.

Knowledge artifacts (prompts, agent skills, patterns, documentation) are licensed under CC BY-SA 4.0. As pure knowledge, they should remain free permanently — and anyone who builds on them should extend the same freedom. Attribution keeps the chain of origin visible.

Libraries are licensed under LGPL-3.0. The source code — the knowledge — stays free. You can build whatever you want with it, but modifications to the library itself return to the commons.

Standalone applications and tools are licensed under GPL-3.0 or AGPL-3.0, depending on how the software reaches its users. Tools that run locally use GPL-3.0 — modifications must be shared when the modified software is distributed. Tools deployed as network services use AGPL-3.0 — because serving software over a network doesn’t change the nature of the knowledge it contains. In both cases, modifications to source code are new knowledge, and knowledge belongs to everyone.

On AI

We are not anti-AI. We are anti-extraction. AI should help people understand, build, and share software — but without weakening workers, obscuring ownership, or stealing the commons.

We believe AI can be used responsibly to deepen understanding, remove drudgery, and widen participation — but that careless adoption can just as easily replace understanding, create a maintenance nightmare, and cheapen labor.

The growth of AI comes with high costs — energy, water, hardware, land, emissions — and these costs are distributed across communities who had no say in the matter. We use AI deliberately, and we do not burn compute to generate noise.