skill-usage-log-reviewer

Pass

Audited by Gen Agent Trust Hub on May 15, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill performs as described, focusing on analyzing usage metrics and skill descriptions to provide a deactivation checklist. No malicious patterns were identified in the instructions or documentation.
  • [DATA_EXFILTRATION]: There are no network requests or data transmission patterns. The skill processes user-provided skill lists and internal platform context locally to generate its report.
  • [COMMAND_EXECUTION]: No shell commands or script execution patterns were found. The skill mentions writing a report to the workspace if file system access is available, which is standard reporting behavior.
  • [EXTERNAL_DOWNLOADS]: The skill does not download any external packages, scripts, or configurations. It has no external dependencies.
  • [PROMPT_INJECTION]: The skill's instructions do not contain any attempt to bypass security filters, ignore previous instructions, or extract system prompts. It includes logic to validate that input data is actually a skill list rather than unrelated conversation.
Audit Metadata
Risk Level
SAFE
Analyzed
May 15, 2026, 09:40 AM
Security Audit — agent-trust-hub — skill-usage-log-reviewer