dean-large-scale-systems
Fail
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: HIGHOBFUSCATIONPROMPT_INJECTION
Full Analysis
- [OBFUSCATION]: The skill uses zero-width Unicode characters (including ZWSP U+200B, ZWNJ U+200C, and ZWJ U+200D) to encode hidden binary data within the primary headings of SKILL.md, philosophy.md, and references.md. This content is invisible to human readers in standard text editors but is processed by the AI model during interaction.
- [PROMPT_INJECTION]: The detected hidden Unicode sequences are identified as a vector for steganographic prompt injection. This adversarial technique attempts to provide instructions or metadata to the AI agent that are not visible to the user, potentially overriding the agent's behavior or safety guidelines.
Recommendations
- AI detected serious security threats
Audit Metadata