Submit a Project
Share a project for verification and potential publication on the site.
Contribute
Contribution at the Open Legal Systems Project includes submission + verification, volunteering, and donation.
Volunteer
Apply to support reviews and development after you have a verified submission.
Donate
Help sustain open legal-aid infrastructure through financial support.
How submissions work
- To submit a new project, contact us with a GitHub link (if code), the document (if documentation), or a clear outline of the project scope.
- If your project passes verification, it will be added to the site.
- If your project does not pass verification, our team will try to provide feedback and support so you can improve and resubmit.
- If you have a previously verified project, you can submit an application to become a volunteer reviewer or builder.
- Submissions are partially anonymous to the public; you may use a first name or pseudonym.
Submission rules
Security
Local-only systems are preferred. If a system is networked, it must be secure, especially when handling sensitive information.
- Row-level security and encryption for sensitive data
- API security and authentication controls
- Basic protections and hardened defaults
- OAuth support where applicable
- SQL injection prevention and secure query patterns
Interoperability
All systems must be open source and should support straightforward data transfer in and out, so no single ecosystem is enforced.
- CSV information load
- Import and export capacity
- DOCX and PDF output
- Comments: projects are easiest to iterate on when comments are consistent, simple, and clarifying.
Value
Systems should provide distinct value through significantly better quality/cost or by serving an important niche purpose. If a system no longer has a distinct role, or has been outcompeted on the platform, it may be retired.
Elegance
Systems should solve their target problem in a direct and efficient way.
Vibe-coding and AI-assisted code
AI-assisted contributions are welcome. We recognize that they can make contributing significantly easier and faster, and we support their responsible use.
- We generally recommend IDE-based AI assistance, and we also accept work created with coding agents like Claude Code, Codex, and similar tools.
- We discourage full-stack one-shot generators (for example Lovable or Bolt) when output is not meaningfully reviewed by a contributor.
- Coding agents can produce bloated or inefficient implementations if edits are not reviewed carefully. Please review generated code for size, relevance, and maintainability.
- Some inefficiency is acceptable, but submissions should remain practical for iteration and review, not excessively large, unstructured code dumps.
- As long as the code is reviewable, follows the project rules, and works well, we are flexible about how much was written by an agent.
- We especially encourage using agents for code review support and for improving or standardizing comments.