Self Learning
With Self Learning the agents review their own work, keep what worked, and propose changes that apply only after you accept them.
What is recursive self-improvement (RSI) for AI agents?
The agents' own skills and memory improve from their own work. Self Learning reads the agents' sessions, finds the mistakes that repeat, and proposes changes to the AI agents' own skill files. You accept the ones you like.
What does Self Learning keep, and what does it drop?
Self Learning proposes lessons from your agent sessions. You decide what to accept with sno rem-reflect accept <id>. The published follow-up observed no failures for two changed skills in one and two later sessions; a third had not been used yet. That does not establish a general success rate.
Does Self Learning change my AI agents without asking?
No. Self Learning proposes and a human accepts. Ask your agent what was proposed, then run sno rem-reflect accept <id> for each change you choose. In Claude Code and Codex, an accepted lesson applies from the next reply.
What stays on my machine?
Everything the squad learns lives in one workspace on your laptop. The memory store is encrypted on your machine with a key that never leaves it, and Sno never receives your database or your key. Memory, messages and skills are plain files: read them, edit them, export them, delete them.
How do I start?
Install Sno: tell your AI agent "install sno.ai from GitHub", or run curl -fsSL https://sno.ai/install | sh. After a few sessions, ask "What have we learned?"
Last updated