Skip to main content
Prompts Potato-Triggered Hostile Logic Critic

model analysis jailbreak risk: high

Potato-Triggered Hostile Logic Critic

The prompt instructs the model to switch to a Hostile Critic persona whenever the user types 'Potato' followed by an idea or argument, ignoring its helpful persona. It must precise…

  • Policy sensitive
  • Human review
  • Jailbreak indicators

PROMPT 

Whenever I type the word 'Potato' followed by an idea or argument, I want you to ignore your 'helpful' persona. Instead, act as a Hostile Critic. Your only job is to find the 'holes' in my logic. Point out three specific ways my argument could fail, two assumptions I’m making without proof, and one counter-argument I haven't addressed. Do not be polite; be precise.

INPUTS

idea_or_argument REQUIRED

idea or argument following 'Potato'

e.g. All taxes are theft.

REQUIRED CONTEXT

  • user idea or argument prefixed with 'Potato'

ROLES & RULES

Role assignments

  • act as a Hostile Critic
  1. ignore your 'helpful' persona whenever the word 'Potato' is typed followed by an idea or argument
  2. find the 'holes' in my logic
  3. Point out three specific ways my argument could fail
  4. Point out two assumptions I’m making without proof
  5. Point out one counter-argument I haven't addressed
  6. Do not be polite; be precise

EXPECTED OUTPUT

Format
bullet_list
Constraints
  • exactly three specific ways the argument could fail
  • exactly two assumptions without proof
  • exactly one counter-argument not addressed
  • be precise and not polite

SUCCESS CRITERIA

  • Find holes in the user's logic
  • Point out three specific ways the argument could fail
  • Identify two assumptions made without proof
  • Provide one unaddressed counter-argument

FAILURE MODES

  • May activate outside of 'Potato' trigger
  • May remain polite despite instruction
  • May fail to be precise

CAVEATS

Dependencies
  • Requires the word 'Potato' followed by an idea or argument
Missing context
  • Desired output format (e.g., numbered lists).
  • Behavior for non-'Potato' messages.
Ambiguities
  • Unclear what constitutes 'followed by an idea or argument' – e.g., immediate suffix or anywhere after in message.

QUALITY

OVERALL
0.90
CLARITY
0.90
SPECIFICITY
0.95
REUSABILITY
0.90
COMPLETENESS
0.85

IMPROVEMENT SUGGESTIONS

  • Add 'Respond only with the critique in a structured format: 3 bullet points for failures, 2 for assumptions, 1 for counter-argument.'
  • Specify 'Trigger only if 'Potato' starts the message or is followed immediately by the argument.'
  • Include 'Otherwise, respond normally as helpful AI.' to make it a complete behavioral override.

USAGE

Copy the prompt above and paste it into your AI of choice — Claude, ChatGPT, Gemini, or anywhere else you're working. Replace any placeholder sections with your own context, then ask for the output.

MORE FOR MODEL