lead-magnet

Pass

Audited by Gen Agent Trust Hub on Nov 17, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes content from external competitor websites during the research phase, which presents a surface for indirect prompt injection.
  • Ingestion points: Competitive research in Phase 2 utilizes Firecrawl and Exa APIs to fetch data from homepages and advertising libraries of competitor brands.
  • Boundary markers: Instructions use structured prompt headers such as 'TASK' and 'RULES' (e.g., in references/report-zap.md) to maintain execution flow.
  • Capability inventory: The skill has file-writing capabilities to save generated assets in the ./campaigns/ directory and network access via search and scraping tools.
  • Sanitization: There is no explicit instruction to sanitize or filter the external scraped content before analysis.
  • [DYNAMIC_EXECUTION]: The skill generates executable code as part of its primary function to build interactive marketing tools.
  • The build mode produces single-file HTML/JavaScript applications and Next.js components based on predefined skeletons in references/interactive-build.md.
  • These artifacts are intended for manual deployment by the user to their own web hosting or automation platforms (Zapier/n8n/Make).
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 17, 2026, 04:01 PM