brand-copywriter
Pass
Audited by Gen Agent Trust Hub on Mar 25, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection during the voice discovery workflow.
- Ingestion points: Instructions in
SKILL.mdandreferences/brand-voice-learning.mddirect the agent to gather and analyze 3-5 external brand assets (e.g., homepages, emails, ads) provided by the user via text or links. - Boundary markers: Absent. There are no instructions or delimiters defined to prevent the agent from following commands that might be embedded within the brand assets being analyzed.
- Capability inventory: The skill performs file read/write operations (e.g., creating
brand-voice-profile.md) and generates natural language copy. It does not possess network access, shell execution, or subprocess capabilities, which significantly limits the impact of a successful injection. - Sanitization: Absent. The skill does not implement validation or filtering on the content of the ingested brand assets.
Audit Metadata