compliance-review

Pass

Audited by Gen Agent Trust Hub on Apr 19, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill instructions and phases are consistent with the stated purpose of technical compliance auditing and follow industry standards.
  • [PROMPT_INJECTION]: The skill demonstrates a surface for indirect prompt injection as it is designed to ingest and process data from external target URLs. Ingestion points: Target URLs and browser console logs are processed in Phase 3. Boundary markers: None identified; untrusted web content is processed without explicit delimiters. Capability inventory: The skill uses browser interaction tools and local file writing for reports. Sanitization: No sanitization of external web content is specified.
  • [DATA_EXFILTRATION]: All data processing is contained within the specified scope, and reports are written to vendor-controlled local directories (shiplight/), posing no exfiltration risk.
  • [COMMAND_EXECUTION]: While the skill generates YAML-based regression tests for continuous compliance, these are data-driven configurations and do not involve arbitrary shell command execution.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 19, 2026, 06:04 PM