skill-manager

Pass

Audited by Gen Agent Trust Hub on Apr 2, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill provides instructions for the agent to run a local utility script (~/.claude/scripts/skill-manager.sh) to perform system maintenance and reporting.
  • [EXTERNAL_DOWNLOADS]: The skill references checking for updates using npx, which connects to the official npm registry to identify newer versions of packages.
  • [DATA_EXFILTRATION]: The skill accesses a local data file (~/.claude/skill-usage.jsonl) to analyze and report on skill invocation frequency based on internal metadata.
  • [PROMPT_INJECTION]: The skill performs audits on other skills by reading their descriptions for duplicate detection, establishing an indirect data ingestion surface. 1. Ingestion points: Reads metadata from files in the local skill repository. 2. Boundary markers: None identified. 3. Capability inventory: Execution of local shell scripts. 4. Sanitization: Not explicitly documented for description parsing.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 2, 2026, 07:19 PM