full-output-enforcement
Pass
Audited by Gen Agent Trust Hub on Aug 11, 2026
Risk Level: SAFEPROMPT_INJECTIONNO_CODE
Full Analysis
- [PROMPT_INJECTION]: The skill uses assertive and overriding instructions (e.g., "No exceptions," "hard failures," and "Overrides default LLM truncation behavior") to steer the agent's internal logic. This technique is intended to bypass the model's default constraints on output length and brevity in favor of exhaustive generation.\n- [NO_CODE]: The skill does not contain any executable scripts, tool definitions, or binary files. It relies entirely on natural language instructions to influence agent behavior.
Audit Metadata