johnny-suede-write
Pass
Audited by Gen Agent Trust Hub on Oct 13, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill acts as a text-processing utility that ingests and transforms untrusted user content and local documentation.
- Ingestion points: Processes user-supplied copy, as well as project files such as
PRODUCT.md,README.md, andAGENTS.md(as specified inSKILL.md). - Boundary markers: Employs an 'Evidence Boundaries' protocol and 'Slop Stop' pass to distinguish between creative rewriting and factual verification.
- Capability inventory: The skill is strictly limited to text manipulation; no instances of shell command execution, file system writes, or network operations were found in the instructions.
- Sanitization: Includes a multi-dimensional scoring system (70-point score) and a 'Ship Gate' to ensure output quality and factual accuracy.
Audit Metadata