skill-manager

Pass

Audited by Gen Agent Trust Hub on Apr 1, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses common shell utilities such as ls, grep, mkdir, and cat. These commands are restricted to the agent's configuration directory (~/.claude/skills/) and are necessary for the skill's primary function of managing instruction files.
  • [PROMPT_INJECTION]: The skill provides instructions for the agent to autonomously identify and save new knowledge. These are functional guidelines intended to improve agent performance over time and do not contain instructions to bypass safety filters, ignore system prompts, or escalate privileges.
  • [DATA_EXFILTRATION]: No network activity or external data transfers were detected. The skill includes a proactive quality checklist that explicitly directs the agent to ensure no credentials or sensitive internal URLs are saved in the generated skills.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 1, 2026, 05:03 PM