razzant/ouroboros

What it solves

Ouroboros is a general-purpose AI agent designed to maintain a consistent identity, durable memory, and history across tasks and restarts. It addresses the limitation of stateless AI agents by allowing the agent to evolve its own implementation—including its code, architecture, prompts, and tools—while maintaining continuity of self-understanding and biography.

How it works

Ouroboros operates as a native desktop application or a headless CLI. It manages its own repository, memory, and history locally on the user's machine. For cognitive processing, it can connect to remote APIs or use local GGUF models. It utilizes a "reviewed core-evolution system" where improvements are turned into reviewed changes in its Git history. To handle coding tasks, it bundles Claudexor as a local execution layer to drive existing coding subscriptions like Codex, Claude Code, or Cursor.

Who it’s for

It is designed for users who need a long-term AI collaborator that can work on external projects, coordinate swarms of specialist agents, and autonomously evolve its own capabilities over time.

Highlights

• Self-Modification: Can rewrite its own application code, architecture, prompts, and dependencies.
• Durable Continuity: Maintains identity, memory, and dialogue history across restarts.
• Swarm Coordination: Can deploy and coordinate a live swarm of specialist agents to work in parallel.
• Background Consciousness: Supports reflection and initiative outside of the immediate request-response loop.
• Cross-Platform Support: Available as a native app for macOS, Windows, and various Linux distributions.
• Extensibility: Expandable via reviewed skills, transport bridges, and MCP tools through OuroborosHub.