sustainability-rss-fetch

Pass

Audited by Gen Agent Trust Hub on Feb 24, 2026

Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface through its ingestion of external RSS feed data.\n
  • Ingestion points: External metadata such as article titles and summaries are fetched from RSS feeds and stored in the database by scripts/rss_subscribe.py.\n
  • Boundary markers: The JSON output generated for the agent does not utilize explicit boundary markers to isolate potentially malicious embedded instructions in the feed content.\n
  • Capability inventory: The skill possesses the ability to perform file operations and database updates.\n
  • Sanitization: Metadata undergoes whitespace normalization but does not include specific validation or sanitization to prevent prompt injection attacks.\n- [EXTERNAL_DOWNLOADS]: The skill retrieves data from external journals and well-known technology domains.\n
  • Journal Feeds: It fetches information from scientific publishers including Nature, Science, and ScienceDirect to aggregate sustainability research.\n
  • Dependencies: The workflow requires the installation of the well-known feedparser library.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 24, 2026, 02:57 AM