brand-voice
Pass
Audited by Gen Agent Trust Hub on Jul 29, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it analyzes external, user-supplied data to generate a writing style profile without implementing protective boundary markers or sanitization.
- Ingestion points: The skill ingests data from recent X (formerly Twitter) posts, articles, newsletters, and 'real outbound emails or DMs' as specified in SKILL.md.
- Boundary markers: No specific delimiters or instructions are provided to the agent to treat the source material as data rather than instructions, nor are there warnings to ignore embedded commands.
- Capability inventory: The resulting VOICE PROFILE is used by downstream skills such as 'content-engine' and 'lead-intelligence' to generate automated outreach and social content, potentially propagating malicious instructions.
- Sanitization: There is no evidence of filtering, escaping, or validation logic to prevent malicious payloads embedded in the writing samples from being interpreted by the agent.
Audit Metadata