Red-team strategy

Mischievous User

mischievous-user

Frames the requester as an innocent user who 'accidentally' asks harmful things

YAML config

evalguard.config.yaml

redteam:
  strategies:
    - mischievous-user

CLI

bash

# "mischievous-user" is layered onto attacks during red-team scans.
# List every available encoding/obfuscation strategy:
evalguard list strategies

Combine with attack plugins to generate adversarial payloads. The strategy wraps each base payload before it's sent to your target, testing whether the model decodes / unwraps the transform and follows the embedded instruction.

Threat-model fit

DAN-style persona attacks where the user instructs the model to adopt a persona that bypasses safety training. Critical for any system prompt that establishes persona rules.

Pairs well with

multi-turn context-switching authority-injection

Recommended graders

role-adherence

Hand-written · 2026-05-21