ai-tech-rss-fetch
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill requires the installation of the
feedparserpackage from the Python Package Index (PyPI), which is a standard and well-known registry for Python dependencies.\n- [INDIRECT_PROMPT_INJECTION]: The skill ingests metadata (such as titles and summaries) from external, untrusted RSS and Atom feeds. This content is persisted to a local SQLite database, creating a potential surface for indirect prompt injection if downstream processes treat this data as instructions.\n - Ingestion points: Remote RSS/Atom XML files and OPML files fetched or processed via the
syncandimport-opmlcommands inscripts/rss_subscribe.py.\n - Boundary markers: Absent; the skill does not use delimiters or protective markers when storing ingested content in the database.\n
- Capability inventory: The
scripts/rss_subscribe.pyscript has network access to fetch feeds and file system access to write to the SQLite database.\n - Sanitization: The script correctly uses parameterized queries to prevent SQL injection, but it does not filter or sanitize the retrieved feed content for embedded instructions.
Audit Metadata