blogwatcher

Pass

Audited by Gen Agent Trust Hub on Mar 1, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill processes untrusted data from external RSS/Atom feeds, creating a surface for indirect prompt injection if an agent reads the resulting notifications.\n- Ingestion points: The fetchFeed function in blogwatcher.js retrieves XML content from arbitrary URLs defined in the feeds configuration.\n- Boundary markers: The skill does not implement delimiters or instructions for downstream agents to ignore embedded commands within the feed content.\n- Capability inventory: The skill has capabilities for file system access (fs.writeFileSync in saveSeenEntries) and network operations (http.get, https.get, and fetch).\n- Sanitization: While the skill truncates text for display, it does not perform sanitization of the feed content (such as stripping instructional keywords) before outputting it.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 1, 2026, 05:11 AM