clarify-unknown

Pass

Audited by Gen Agent Trust Hub on Apr 17, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill correctly implements its strategic analysis function. Its operations, including context gathering and file writing, are legitimate and handled through standard agent tools.
  • [INDIRECT_PROMPT_INJECTION]: The skill has a surface for indirect prompt injection through its ingestion of external files. Ingestion points: Analysis strategy documents and project files (README, CLAUDE.md). Boundary markers: Not explicitly defined. Capability inventory: File reading, file writing, and user questioning tool (AskUserQuestion). Sanitization: No explicit sanitization or filtering logic is present.
  • [DATA_EXPOSURE_AND_EXFILTRATION]: The skill accesses project-level files for context. It does not access sensitive credential files, and there are no network operations or instructions to exfiltrate data.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 17, 2026, 06:07 AM