skill-auditor

Pass

Audited by Gen Agent Trust Hub on May 18, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: Static analysis hints regarding prompt injection in references/threat-patterns.md and scripts/analyzers/static.js are false positives. These files contain the detection logic (regex patterns and documentation) used by the auditor to identify attacks in other skills, rather than being injection attempts themselves.
  • [COMMAND_EXECUTION]: The skill utilizes Node.js child_process and Python subprocess modules to execute analysis scripts and dataflow trackers. This functionality is essential for its primary purpose as an auditing tool and is fully disclosed.
  • [EXTERNAL_DOWNLOADS]: The setup wizard (scripts/setup.js) offers an opt-in feature to install legitimate security analysis packages (tree-sitter, tree-sitter-python) from official registries via pip. These are well-known developer tools.
  • [SAFE]: Network operations in scripts/scan-url.js and scripts/analyzers/virustotal.js are used to fetch remote skill content for scanning and check binary reputations, respectively. These actions align with the tool's stated security mission.
  • [SAFE]: The skill accesses the file system to read local skill directories provided by the user and saves its own configuration to ~/.openclaw/skill-auditor.json, which is standard behavior for local CLI utilities.
Audit Metadata
Risk Level
SAFE
Analyzed
May 18, 2026, 01:02 AM
Security Audit — agent-trust-hub — skill-auditor