brand-voice
Pass
Audited by Gen Agent Trust Hub on Sep 6, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONNO_CODE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process external content, specifically user-provided text drafts and a configuration file (references/brand-voice.md). This creates a surface for indirect prompt injection where instructions hidden in the processed data could attempt to steer the agent's behavior. 1. Ingestion points: references/brand-voice.md and user-supplied drafts. 2. Boundary markers: None. 3. Capability inventory: Text rewriting only as described in SKILL.md; no tools, network, or file system access detected. 4. Sanitization: None.
- [NO_CODE]: The skill consists entirely of instructional markdown and configuration templates without any executable scripts or binary files.
- [SAFE]: The skill's behavior is restricted to text transformation within the conversation context, with no evidence of malicious intent or dangerous system interactions.
Audit Metadata