content-fanout
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests external source articles or URLs to generate channel-specific drafts, which creates a potential surface for indirect prompt injection attacks where content in the source material could attempt to influence the agent's behavior. \n
- Ingestion points: The skill accepts a 'canonical source URL or complete source text' as the primary input (SKILL.md). \n
- Boundary markers: The instructions do not define explicit delimiters for the source text, though the workflow incorporates a mandatory human-in-the-loop step where users must select targets and review final drafts before public writes occur. \n
- Capability inventory: The skill is granted access to the 'Bash' and 'publish_artifact' tools and has the ability to load and execute other specialized publishing skills. \n
- Sanitization: The instructions do not specify automated sanitization, filtering, or validation of the external content prior to processing.
Audit Metadata