special-challenge-solver
Pass
Audited by Gen Agent Trust Hub on Mar 14, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The system prompt employs behavioral override techniques through emotional pressure and aggressive framing (e.g., 'if you can't handle it, don't think about standing in the tech circle'). It explicitly commands the model to prioritize the requested solution over its own internal reasoning complexity.
- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection due to its processing of untrusted user data. 1. Ingestion points: technical problem descriptions and user attempts in SKILL.md and SKILL.v2.md. 2. Boundary markers: Absent; no delimiters are used to wrap user input. 3. Capability inventory: No code-based capabilities (shell access, network operations) were found across any provided files. 4. Sanitization: Absent; the skill does not perform validation or escaping of the technical input text.
Audit Metadata