Self Improving Agent 1 0 2 Litiao

Captures learnings, errors, and corrections to enable continuous improvement. Use when: (1) A command or operation fails unexpectedly, (2) User corrects Clau...

This package appears to do what it says: reminders, error-detection based on tool output, and a helper to scaffold extracted skills. Before enabling it: (1) review the three scripts (activator.sh, error-detector.sh, extract-skill.sh) to ensure you are comfortable with file paths and behavior; (2) prefer adding hooks at the project-level rather than user-level if you only want it active per-repo — user-level settings make the hooks run globally; (3) be aware error-detector relies on the CLAUDE_TOOL_OUTPUT env var (provided by the host) and extract-skill.sh will create files under ./skills or wherever SKILLS_DIR points — confirm that location and permissions are acceptable; (4) no credentials are requested, and there are no outbound network calls in the scripts, but if you later modify scripts to send or publish learnings, re-review for exfiltration risks.

Type: OpenClaw Skill Name: self-improving-agent-1-0-2-litiao Version: 1.0.0 The bundle is a legitimate productivity tool designed to help AI agents log errors, user corrections, and new knowledge into markdown files for continuous improvement. It includes shell scripts (activator.sh, error-detector.sh, extract-skill.sh) that provide context reminders, detect command failures via environment variables, and scaffold new skill directories. The logic is transparent, lacks any data exfiltration or remote execution capabilities, and includes input validation in the extraction script to prevent path traversal.