competitor-intel
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill operates as an integration layer that ingests external web data, establishing an inherent indirect prompt injection risk profile.
- Ingestion points: Untrusted content enters the agent's context from competitor websites, landing pages, and public social feeds via tools like
firecrawl_urls_scrape,web_pages_scrape, and platform-specific scrapers as described inSKILL.mdandreferences/source-by-source.md. - Boundary markers: Explicit boundary markers or instruction-isolation delimiters are absent in the prompt interpolation process.
- Capability inventory: The skill leverages an inventory of web-scraping, search, brand-tracking, and image-generation toolkits (
firecrawl_*,hyperseo_*, and various social scrapers). - Sanitization: No custom logic or instructions are present to sanitize, strip, or escape potential prompt injection strings embedded within the retrieved website text before processing.
Audit Metadata