blogwatcher

Pass

Audited by Gen Agent Trust Hub on Oct 5, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and summarize external RSS feeds and web content, which are untrusted sources.
  • Ingestion points: External content is fetched via RSS URLs in SKILL.md using the feedparser library.
  • Boundary markers: The skill's instructions do not specify the use of delimiters or 'ignore' instructions when passing retrieved content to the AI for summarization.
  • Capability inventory: The skill demonstrates the capability to write processed data to the local file system (e.g., feed_digest.json).
  • Sanitization: The provided code snippets do not implement sanitization or filtering to remove potential instructions or scripts from the feed data before processing.
  • [EXTERNAL_DOWNLOADS]: The skill requires the installation of an external Python dependency.
  • Evidence: The documentation instructs the user to run pip install feedparser to enable RSS parsing capabilities.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 5, 2026, 08:06 AM