voice-enforcement
Pass
Audited by Gen Agent Trust Hub on Oct 6, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONNO_CODE
Full Analysis
- [NO_CODE]: The skill does not contain any executable scripts, binaries, or source code files; it consists entirely of Markdown documentation and instructions.
- [COMMAND_EXECUTION]: The skill instructions provide usage examples for an external CLI tool named voice-scorer (e.g., voice-scorer score path/to/doc.md) for document analysis and diffing.
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to have the agent process and review user-provided documents, creating an attack surface for indirect prompt injection through instructions embedded in the analyzed content. * Ingestion points: External prose-heavy documents such as READMEs, design docs, and essays as specified in the skill triggers and description. * Boundary markers: The instructions do not define any specific delimiters or boundary markers to isolate user-provided content from agent instructions. * Capability inventory: The agent is empowered to perform text generation, document review, and to execute the referenced voice-scorer CLI tool. * Sanitization: There are no provisions for sanitizing, escaping, or filtering the content of the documents before the agent processes them.
Audit Metadata