writing-and-editing
Pass
Audited by Gen Agent Trust Hub on Oct 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process user-provided documents and context gathered from external sources (e.g., Google Drive, Slack, Microsoft Teams, SharePoint), which represents a vulnerability surface for indirect prompt injection.
- Ingestion points: Stage 1 context gathering and Stage 3 reader testing involve processing untrusted content from user-pasted text and integrated cloud or messaging services.
- Boundary markers: The instructions do not explicitly mandate the use of robust delimiters (such as XML tags or triple-backticks) or specific "ignore embedded instructions" headers when interpolating untrusted content into agent prompts or sub-agent invocations.
- Capability inventory: The skill's capabilities include creating and modifying markdown files (as platform artifacts or local files), reading from integrated services, and spawning sub-agents for document testing.
- Sanitization: No explicit validation, sanitization, or filtering of ingested external content is mentioned in the provided workflow.
Audit Metadata