COMMUNITY / PROJECTS AND RESEARCH
Ouroboros in projects and research.
Other projects are adapting Ouroboros code and ideas, and research on self-improving agents cites it. Each entry below links to the work itself.
Projects adapting code and ideas
- DPC Messenger adapts Ouroboros's agent for its messaging system, with its own model manager, context, and local state. Its agent source credits Ouroboros and explains the changes.
- offload-harness is porting Ouroboros ideas. Its proposed design decision lists a failure-class register, review evidence bound to a commit with an explicit NOT_RUN outcome, and deterministic gates before model review. Ouroboros itself stays parked there.
Research that cites Ouroboros
- Self-Evolving Coding Agents discusses persistent implementation changes and the distinction between a frozen benchmark snapshot and an agent that keeps developing.
- The Last AI Built by Humans: Toward Genuine Recursive Self-Improvement discusses Ouroboros as a case of post-deployment changes to agent software.
- RSIAgent: Autonomous Exploration for Recursive Self-improvement in New Environments cites Ouroboros in its related work on recursive self-improvement.
- MetaRSI / RSI2: A Meta-Recursive Self-Improving System for Recursive Self-Improving Systems Themselves includes Ouroboros in its survey of evaluation domains and verification methods.
Share what you are building
If you have adapted Ouroboros, used its public evidence, or written about it, tell us in Discussions. Contributions to Ouroboros itself start with the contribution guide.