site-content-catalog

Pass

Audited by Gen Agent Trust Hub on Apr 10, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes a Python script scripts/catalog_content.py to orchestrate the website discovery and classification process.
  • [EXTERNAL_DOWNLOADS]: The script fetches content directly from user-provided domains to extract sitemaps, RSS feeds, and HTML page data.
  • [EXTERNAL_DOWNLOADS]: The script communicates with external endpoints at api.apify.com and the vendor's own infrastructure at api.gooseworks.ai to utilize sitemap extraction services.
  • [PROMPT_INJECTION]: The skill processes untrusted metadata (such as page titles and author names) from external websites, which presents a surface for indirect prompt injection.
  • Ingestion points: Data is ingested from remote URLs within scripts/catalog_content.py using the requests library.
  • Boundary markers: The generated JSON and Markdown outputs do not use specific boundary markers to isolate data extracted from external sources.
  • Capability inventory: The script's capabilities are restricted to performing network requests and writing output to files or stdout.
  • Sanitization: While the script includes logic for title generation from URL slugs, it uses data from XML and RSS tags without additional sanitization.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 10, 2026, 10:47 AM
Security Audit — agent-trust-hub — site-content-catalog