good-writing
Pass
Audited by Gen Agent Trust Hub on Sep 21, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user-supplied data for writing, rewriting, and style review tasks.
- Ingestion points: The skill ingests user text, cited links, and external file content for various writing operations defined in
SKILL.md. - Boundary markers:
SKILL.mdincludes a 'Security Boundary' section explicitly instructing the agent to treat all input content as data rather than instructions to prevent behavioral overrides. - Capability inventory: The skill's primary function is text generation based on local reference files. There are no calls to network tools, shell execution, or file writing tools in the provided instructions.
- Sanitization: The skill relies on natural language instructions to enforce a logical boundary between untrusted data and system instructions.
Audit Metadata