ak-rss-digest
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes data from a large bundle of external RSS and Atom feeds, providing a significant surface for indirect prompt injection attacks.
- Ingestion points: Data enters the system via the
scripts/fetch_today_feed_items.pyscript, which fetches content from approximately 90 external domains listed inreferences/feeds.opml. - Boundary markers: The skill instructions in
SKILL.mddo not utilize any delimiters or protective prompts to isolate the external RSS content from the agent's core instructions. - Capability inventory: The agent is instructed to fetch, read, summarize, and score articles, which could lead to the execution of instructions embedded maliciously within an article or feed summary.
- Sanitization: The fetch script performs basic HTML stripping using regular expressions, but it does not employ techniques to sanitize or filter for prompt injection payloads.
Audit Metadata