Self Improving Agent

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: add lightweight reminders and local logs for learnings/errors. Before installing or enabling hooks: 1) Review the hook scripts (activator.sh, error-detector.sh, handler.js/ts) to confirm you are comfortable with them running in your environment. 2) Do not enable the PostToolUse hook (error-detector) globally unless you trust that tool outputs won't contain secrets — the detector reads CLAUDE_TOOL_OUTPUT and could cause you (or the agent) to copy sensitive output into .learnings if you ignore the skill's redaction guidance. 3) When following the manual install steps, verify the GitHub repo URL and author; cloning from an external repo is expected but worth confirming. 4) If you enable hooks by copying into ~/.openclaw/hooks, remember that hook scripts run with your user permissions — enable only on systems you control. 5) Test extract-skill.sh with --dry-run first; the script enforces relative output paths. If you want extra assurance, run the hooks in a disposable workspace or inspect behavior with low-privilege accounts before enabling in production.

Type: OpenClaw Skill Name: self-improving-agent-grant Version: 3.0.16 The self-improving-agent-grant skill bundle is a productivity tool designed to help AI agents maintain a local log of errors, corrections, and feature requests in a `.learnings/` directory. It utilizes OpenClaw hooks (`hooks/openclaw/handler.js`) to inject reminders into the agent's bootstrap process and provides shell scripts (`scripts/extract-skill.sh`, `scripts/error-detector.sh`) to automate the management of these logs. The code demonstrates defensive practices, such as explicit instructions to avoid logging secrets and path validation in scripts to prevent directory traversal, and shows no evidence of malicious intent or data exfiltration.