writing-shape
Pass
Audited by Gen Agent Trust Hub on Sep 24, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill operates as a document structuring tool, focusing on reading and organizing text. It does not utilize network tools, execute code, or access sensitive system directories. Instructions emphasize that the raw material file is read-only.
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted material from a source file ('the pile'). While this creates an injection surface, the risk is mitigated by logic requiring the agent to paraphrase the content and collaborate with the user on specific structure.
- Ingestion points: Raw material markdown file (SKILL.md).
- Boundary markers: Absent.
- Capability inventory: Local file read and file append operations.
- Sanitization: Content is reworked and paraphrased into a user-verified article structure rather than being interpreted as instructions.
Audit Metadata