RSIm — Recursively Self-Improving Module
First stable release.
What's included
- 30+ skills: cycle / system / security / domain / design / deploy
- Improvement cycle:
/reflect→/improve(L1 knowledge / L2 workflow) - Four memory layers: episodic failures, heuristics, semantic patterns, procedural skills
- Export & sync:
/rsim-export(contribute via PR) +/rsim-sync(pull updates) - Documentation:
docs/rsim-for-ai.md,docs/rsim-guide.md,docs/rsim-paper.md
Install
Unix / macOS / WSL
curl -fsSL https://raw.githubusercontent.com/the-hugin/rsim/main/install.sh | bashWindows PowerShell
irm https://raw.githubusercontent.com/the-hugin/rsim/main/install.ps1 | iexRequirements
- Claude Code
- gh CLI +
gh auth login(for/rsim-exportand/rsim-synconly)