stonebraker-database-architecture
Fail
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: HIGHOBFUSCATIONPROMPT_INJECTION
Full Analysis
- [OBFUSCATION]: The skill header in SKILL.md contains a sequence of zero-width characters (U+200B, U+200C, U+200D, U+2060) that hide an encoded string. This zero-width steganography is a technique used to embed hidden data that is invisible to users but processed by the AI model.
- [PROMPT_INJECTION]: The use of concealed Unicode content to hide instructions within the skill represents a high-severity prompt injection vector. This method is intended to bypass safety mechanisms and human review by providing instructions that the agent can see but the user cannot, potentially leading to unauthorized or malicious actions.
Recommendations
- AI detected serious security threats
Audit Metadata