business-review
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from various external sources, creating a surface for indirect prompt injection attacks where malicious instructions could be embedded in analyzed content.
- Ingestion points: External URLs fetched via
WebFetch, local repository files (READMEs, marketing strings), and user-provided documents (copy decks, transcripts, slide decks). - Boundary markers: The instructions lack explicit delimitation or "ignore embedded instructions" directives when processing these sources.
- Capability inventory: The skill can perform network requests (
WebFetch), read local files, and write new markdown files to thedocs/directory. - Sanitization: No filtering, escaping, or validation of the ingested content is mentioned before it is interpolated into the reasoning process.
Audit Metadata