doc-digest
Pass
Audited by Gen Agent Trust Hub on Oct 26, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process and summarize potentially untrusted external content such as PDFs, transcripts, and logs provided by users.
- Ingestion points: Processes user-provided attachments and pasted text content.
- Boundary markers: Instructions specify wrapping the output in a single code block, which helps separate the summary from the rest of the conversation.
- Capability inventory: The skill does not use any dangerous tools; it has no network access, no file system write permissions, and no shell execution capabilities.
- Sanitization: No explicit sanitization of document content is performed, but the lack of exploitable tools limits the impact of potential injection attacks.
Audit Metadata