user-research

Pass

Audited by Gen Agent Trust Hub on Mar 16, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No malicious code or security threats were identified. The skill is composed of markdown instructional content and documentation templates.\n- [PROMPT_INJECTION]: No direct prompt injection or bypass instructions were found. An analysis of the indirect prompt injection surface was performed due to the presence of web-fetching tools.\n
  • Ingestion points: WebFetch and WebSearch tools defined in SKILL.md\n
  • Boundary markers: Absent in provided research templates\n
  • Capability inventory: Read, Glob, Grep, WebFetch, and WebSearch as listed in SKILL.md\n
  • Sanitization: Not applicable for these static markdown templates\n- [DATA_EXFILTRATION]: No exposure of sensitive file paths, hardcoded credentials, or unauthorized data transmission patterns were detected.\n- [REMOTE_CODE_EXECUTION]: No remote code execution patterns, external script downloads, or package installation commands were found. The skill does not interact with external code sources.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 16, 2026, 02:24 PM