linkedin-post-style
Pass
Audited by Gen Agent Trust Hub on Oct 12, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill is entirely descriptive, providing instructions for text generation and content formatting without any executable scripts or network-bound operations.
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied notes or drafts to generate posts. While it does not define strict boundary markers for this data, the risk is minimal as the skill's capabilities are limited to producing text and invoking authorized visual companion tools. Ingestion points: User-provided source material in
SKILL.md. Boundary markers: Absent. Capability inventory: Text, PDF, Image, and Video generation tools. Sanitization: Absent. - [DYNAMIC_CONTEXT_INJECTION]: No use of the
!commandsyntax or other load-time command execution patterns was detected in the skill instructions. - [OBFUSCATION]: No attempts to hide URLs or commands through encoding or other obfuscation techniques were found.
Audit Metadata