skill-rss-aggregator
Pass
Audited by Gen Agent Trust Hub on Oct 16, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill fetches and processes data from external RSS/Atom feeds which are attacker-controlled sources.
- Ingestion points: External XML content is fetched via
urllib.request.urlopeninscripts/rss_digest.pyand parsed to extract titles and summaries. - Boundary markers: None identified in the prompt interpolation; the content is placed directly into Markdown or JSON outputs.
- Capability inventory: The skill has the capability to write files (
Path(a.output).write_text) and make network requests (urllib.request.urlopen). - Sanitization: Basic HTML stripping is performed via regex (
_strip_html), but there is no specific sanitization against embedded natural language instructions that could influence subsequent agent steps if the output is fed into other skills likeskill-topic-evaluator.
Audit Metadata