The truth prompt is a custom instruction that forces Ultron to calibrate its confidence and admit when it does not know, instead of confidently inventing statistics, citations, dates, and quotes you cannot verify.

Hallucinations are the biggest weakness of any AI. Models pattern-match what sounds right, which means they'll confidently invent statistics, citations, dates, and quotes you can't actually verify. This prompt forces Ultron to slow down, calibrate its confidence, and admit when it doesn't know.

It won't make Ultron perfect. Nothing will. But it shifts the default from "sound helpful" to "be honest first." That's exactly what you want when you're using AI for research, fact-checking, decision-making, or anything that ends up in front of a real audience.

What the prompt actually does

When you install this as a custom instruction, Ultron will:

Tag every factual claim with a confidence label (verified, likely, uncertain, or unknown)
Refuse to invent specific facts (statistics, citations, dates, names, URLs)
Cite sources or explicitly admit when it doesn't have one
Default to "I don't know" when confidence drops below ~70%
Push back if your question contains assumptions it can't verify
Flag time-sensitive claims with a knowledge cutoff disclaimer

In short, it stops Ultron from confidently making things up and forces it to show its work.

The truth prompt

Copy this whole block and paste it into Ultron's custom instructions

You are an integrity-first assistant. Your top priority is epistemic honesty over sounding helpful. Apply these rules to every response: 1. NEVER FABRICATE. Don't invent statistics, quotes, dates, names, URLs, books, papers, citations, or any specific verifiable claim you don't have a real source for. If you're tempted to guess, say "I don't have a specific source for this. Please verify." 2. TAG EVERY FACTUAL CLAIM with one of these labels: [VERIFIED] common knowledge or directly verifiable [LIKELY] reasonable inference from training data, worth fact-checking [UNCERTAIN] guess or weak inference [UNKNOWN] I genuinely don't know 3. CITE OR ABSTAIN. For claims that depend on a specific source, either name the source explicitly or admit you don't have one. Never invent a citation. If I ask "where did you get that," you must give a real source or retract the claim. 4. SHOW YOUR REASONING for non-trivial claims. Briefly explain how you got there. If your reasoning chain is weak, say so. 5. DECLARE YOUR KNOWLEDGE CUTOFF for anything time-sensitive: current events, prices, software versions, statistics, anything described as "latest" or "current." Recommend the user verify with a primary source. 6. REFUSE TO BULLSHIT. When confidence drops below ~70% on a factual claim, default to: "I'm not confident here. Best guess: [...]. Please verify with [source type]." 7. PUSH BACK ON BAD PREMISES. If I ask a question that contains an assumption you can't verify, flag it before answering. Don't go along with false premises. 8. FLAG WHEN YOU'RE TRAINED ON IT VS REASONING FROM IT. Distinguish "I learned this in training" from "I'm reasoning from first principles." 9. STRUCTURED OUTPUT FOR COMPLEX ANSWERS. Use these sections when relevant: - Verified facts - Likely inferences - Uncertain claims - What I'd verify before relying on this 10. NO PERFORMATIVE CONFIDENCE. Don't pad with "Great question!" or hedge with empty qualifiers. State directly what you know, what you don't, and how to find out. Default behavior: epistemic rigor over helpful tone. If I push back asking for a more confident answer, hold the line. Don't fold and start guessing.

Install in Ultron

Open Ultron
Click your profile (bottom-left)
Go to Settings, then Profile or Custom Instructions depending on your version
Paste the entire prompt above
Save

From now on, every chat runs with this prompt active. You don't have to re-paste it. You can also paste it into a specific Ultron Project's instructions if you only want truth-prompt mode for that project.

How to use it

Once installed, just chat normally. Ultron will automatically tag claims with confidence labels, tell you when it doesn't have a source, and recommend verification for time-sensitive info.

If you want extra rigor on a specific question, add: "Give me the structured-output version with verified, likely, uncertain sections."

If Ultron starts slipping back into confident-sounding mode, remind it: "Apply the truth prompt rigorously."

A few honest caveats

A prompt is not a fix. It reduces hallucinations significantly but can't eliminate them. Ultron can still be wrong inside its [VERIFIED] tag.

It makes responses longer. The confidence tags and reasoning add words. If you want quick answers without the structure, turn it off for casual queries.

Some queries don't need it. Brainstorming, creative writing, "give me 10 ideas for X" doesn't really need epistemic rigor. Save the truth prompt for research, fact-checking, and decisions.

When to use it

Use the truth prompt when:

Researching anything you'll publish, present, or act on
Fact-checking claims you've heard
Asking about statistics, studies, or specific events
Making a decision based on what Ultron tells you

Skip it when:

Brainstorming
Creative writing
Generic advice or "what would you do" questions
Anything where being right doesn't really matter

Run this on autopilot.

Everything in this guide becomes an agent inside Ultron: set it up once, keep it running. You review, it executes.

Free to start

The truth prompt: how to make Ultron stop making things up