We build software around real problems, under explicit constraints, in public. No hype cycles. No resume padding. No abandoned repositories.
Projects here are expected to be incomplete, opinionated, and actively worked on.
What we do
OmniForge turns vague problem statements into scoped, buildable systems.
We work on problems that are:
- non-trivial
- constrained by reality
- uncomfortable to finish
Decisions are documented. Tradeoffs are explicit. Failures are kept, not buried.
Shipping and learning matter more here than popularity or growth.
What we don't do
We don't accept idea-only proposals. We don't promise timelines to look busy. We don't run throwaway hackathon projects. We don't optimize for stars, forks, or follower counts.
If you're looking for instant validation, this is not the place.
Active work
OmniForge hosts a small number of active projects at any time.
Each project includes:
- a clear problem definition
- constraints and non-goals
- known limitations
- open issues that actually matter
If nothing looks uncomfortable, we're doing it wrong.
Contributing
There are two ways to engage.
1. Contribute to an existing project
Start with entry-point issues. Expect to read code and documentation before writing anything.
2. Propose a problem worth solving
We accept problem proposals — not solution pitches.
Proposals are reviewed for:
- clarity of the problem
- real-world relevance
- feasibility under open-source constraints
- willingness of the proposer to contribute
Most proposals are rejected. That is intentional.
How decisions are made
Each project has maintainers. Maintainers make final decisions.
Discussion is open. Direction is not a democracy.
This prevents stagnation and keeps projects moving.
Why OmniForge exists
OmniForge exists to build technical judgment.
If something here works, great. If it fails, the failure will be documented. Both outcomes are useful.