Self-Improving Agent
PassAudited by ClawScan on May 13, 2026.
Overview
This instruction-only skill is coherently designed to keep lightweight workspace learning logs, with safeguards against secrets and automatic high-authority memory edits.
Install this if you want the agent to keep local, persistent notes about corrections and workflow lessons. Keep the .learnings/ directory out of shared repos if it may contain private context, and review any proposed promotion into AGENTS.md, MEMORY.md, or similar instruction files before allowing it.
Findings (1)
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
Incorrect, overly broad, or sensitive lessons could persist across sessions or affect future agent behavior if promoted without review.
The skill creates persistent learning files and can promote distilled lessons into long-lived instruction or memory files that may influence future agent behavior.
Create `.learnings/` in the active workspace root or project root... Promote to the smallest durable home: `AGENTS.md` ... `MEMORY.md`
Use this only in workspaces where persistent learning is desired, review .learnings/ entries periodically, and require explicit approval before editing AGENTS.md, MEMORY.md, or project instruction files.