voice-blend
Pass
Audited by Gen Agent Trust Hub on Jun 26, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides instructions and usage examples for executing a local Python script (
voice_blender.py) to process voice profiles. This is part of the core functionality for generating hybrid writing styles. - [SAFE]: Interacts with the file system to read and write voice profiles in standard project (
.aiwg/voices/) and user-level (~/.config/aiwg/voices/) configuration directories. These operations are consistent with the skill's purpose as a profile management tool. - [PROMPT_INJECTION]: The skill demonstrates a surface for indirect prompt injection through the ingestion of external voice profile data from project-local directories.
- Ingestion points: Reads profile definitions from
.aiwg/voices/(SKILL.md). - Boundary markers: No explicit delimiters or instructions to ignore instructions within the profiles are specified in the behavior logic.
- Capability inventory: Performs shell execution of local scripts and file system writes (SKILL.md).
- Sanitization: The skill focuses on numerical interpolation of tone dimensions and merging vocabulary lists without explicitly mentioning sanitization of the strings ingested from those lists.
Audit Metadata