Self Improvement For All
PassAudited by ClawScan on May 1, 2026.
Overview
This skill appears to do what it claims—store AI-agent learnings locally—but those notes persist and can be exported, so avoid recording secrets.
This looks like a benign local learning/memory helper. Before installing, remember that anything the agent records may remain on disk and may be reused in future tasks, so do not store secrets or sensitive internal details unless you are comfortable retaining them locally.
Findings (2)
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.
Sensitive or incorrect notes could be reused in later tasks, leading to repeated mistakes or accidental exposure if the user later exports or shares the data.
This shows the skill creates and reuses persistent task memory. That is the stated purpose, but stored corrections or errors could later influence agent behavior or contain sensitive details.
captures every error, correction, and successful pattern automatically. Then retrieves relevant learnings before tackling similar problems again.
Avoid recording secrets, credentials, or highly sensitive project details; periodically review or delete the local `.adaptive_learning` files.
If the export path is chosen carelessly, a local file could be replaced with the learnings export JSON.
The export function is purpose-aligned, but it writes to a caller-provided path and could overwrite an existing local file if directed there.
def export_learnings(self, output_file: str = "learnings_export.json") ... with open(output_file, 'w') as f:
Export only to intended project or temporary paths, and avoid system, configuration, or credential file locations.