marketing-research

Pass

Audited by Gen Agent Trust Hub on Jul 10, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it processes untrusted web search results.
  • Ingestion points: External data from literary agents, podcasts, and blog listings retrieved via the Perplexity research tool (SKILL.md).
  • Boundary markers: Absent; web content is integrated directly into generated reports without explicit delimiters (SKILL.md).
  • Capability inventory: File-write permissions to save reports to the workspace and research_access for network queries (SKILL.md).
  • Sanitization: No specific input sanitization is documented beyond the use of verified citations provided by the research tool (SKILL.md).
  • [SAFE]: The skill invokes external research tools (Perplexity, OpenRouter) consistent with its role as a marketing research accelerator and vendor resources.
  • [SAFE]: No indicators of malicious code, obfuscation, or unauthorized access were detected in the skill instructions or metadata.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 10, 2026, 03:13 PM
Security Audit — agent-trust-hub — marketing-research