brand-voice
Pass
Audited by Gen Agent Trust Hub on Sep 25, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to ingest and follow instructions from external files and user-provided text samples to define its voice. \n
- Ingestion points: Reads configuration and voice definitions from
CREATOR.md,creator-memory/voice.md,CLAUDE.md, and user-supplied 'Voice Samples'. \n - Boundary markers: The skill lacks explicit instructions for the agent to use delimiters or to ignore potential system-level commands embedded within the voice definitions or samples. \n
- Capability inventory: Primarily used for text generation, draft revision, and terminology enforcement. While the skill itself does not invoke dangerous tools, the content it generates or the logic it adopts from these files influences the agent's overall behavior. \n
- Sanitization: There is no mention of sanitizing or validating the content of the loaded files or samples before they are integrated into the agent's operating context.
Audit Metadata