midwit

Pass

Audited by Gen Agent Trust Hub on Aug 25, 2026

Risk Level: SAFEPROMPT_INJECTION

Full Analysis

[PROMPT_INJECTION]: The skill instructs the agent to persist in a specific persona and behavior pattern for the duration of the session until manually stopped.
[PROMPT_INJECTION]: Instructions explicitly forbid the use of the AskUserQuestion tool, attempting to override standard platform interaction mechanisms.
[PROMPT_INJECTION]: The skill mandates a specific response template for all non-routine interactions, which overrides the agent's standard conversational flow.

Audit Metadata

Risk Level

SAFE

Analyzed

Aug 25, 2026, 03:57 PM

Security Audit — agent-trust-hub — midwit