arabic-market-research

Pass

Audited by Gen Agent Trust Hub on Nov 9, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill instructions do not contain any malicious patterns. It defines a workflow for gathering market evidence and organizing it into content angles without attempting to execute code or access sensitive data.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes external data such as social media comments and competitor channel information. Ingestion points: Step 2 (Gather evidence) involves searching Arabic YouTube, TikTok, Instagram, and community forums like Reddit and Facebook. Boundary markers: The skill includes a specific rule to "Separate evidence from assumption, and label each." Capability inventory: The skill's actions are limited to text analysis and content generation; no subprocess calls, file writing, or network exfiltration tools are requested. Sanitization: No explicit sanitization or filtering of external content is defined. Given that this ingestion is central to the skill's intended purpose and no dangerous capabilities are exposed, this does not pose a significant risk.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 9, 2026, 12:32 PM