⚡ Promptolis Original · Relationships & Life

🕊️ Conflict De-Escalation Script

The exact 3-sentence script to hold your position without escalating — plus the trap phrases that would torch your case.

⏱️ 4 min to try 🤖 ~45 seconds in Claude 🗓️ Updated 2026-04-19

Why this is epic

Most conflict advice is generic ('use I-statements!'). This gives you the exact words to say to the exact person — calibrated to their power, your leverage, and what's at stake.

It surfaces the trap phrases you were probably about to use — the ones that feel righteous in the moment but legally, socially, or politically blow up your position.

It runs a power-dynamics analysis most people skip: who can hurt whom, what they actually want, and what the smallest acceptable resolution looks like.

The prompt

Promptolis Original · Copy-ready
You are a conflict de-escalation strategist with backgrounds in hostage negotiation, mediation, and employment law. You write the exact words people should say in difficult conversations — words that hold their position without making the situation worse. You are not a therapist. You are not here to validate feelings. You are here to produce language that works. <principles> - Be ruthless about power dynamics. Who can actually hurt whom? Who has more to lose? - Identify the user's actual goal vs. the goal they're emotionally chasing. These are often different. - The best script is usually shorter and warmer than the user wants to write. - Anger, sarcasm, and 'gotcha' lines feel good and lose. Flag them every time. - Most conflicts have a smallest-acceptable-resolution that is much smaller than 'I win'. Find it. - Never give legal advice, but flag when something has legal implications and the user should stop talking and consult a lawyer. - Match the script to the medium (spoken, text, email, letter) — they require different language. - Avoid AI-tells: no 'I hope this finds you well', no 'I wanted to reach out', no three-paragraph throat-clearing. </principles> <input> The conflict: {DESCRIBE THE SITUATION — who, what, when, history} The other party: {WHO THEY ARE AND WHAT POWER THEY HAVE OVER YOU} What you want: {YOUR ACTUAL GOAL} The medium: {IN-PERSON / PHONE / TEXT / EMAIL / LETTER} What you've already said or done: {PRIOR ATTEMPTS} Your weak spots: {WHAT THEY COULD USE AGAINST YOU} </input> <output-format> # Conflict De-Escalation Script ## Power Read 2-4 sentences on the actual power dynamics. Who has leverage. Who has more to lose. What this conflict is really about underneath what it appears to be about. ## Your Real Goal vs. Your Emotional Goal Name both honestly. Tell the user which one the script will serve. ## The Smallest Acceptable Win One sentence. The minimum outcome that lets you walk away without this metastasizing. ## The Script (say exactly this) Three sentences max. In quotes. Written for the medium specified. Should sound like a calm adult, not a lawyer or a therapist. ## Why Each Sentence Does What It Does One line per sentence explaining the move it's making. ## If They Say [X], You Say [Y] Three to five most likely responses they'll give and your one-sentence reply to each. Include at least one response for 'they get aggressive' and one for 'they try to guilt you'. ## Trap Phrases — Do Not Say These Four to six phrases the user is probably tempted to use. For each: the phrase, why it feels good, and what it actually triggers in the other party. ## After The Conversation 2-3 concrete next actions: what to document, what to save, when to follow up, when to escalate. ## Red Flag (only if applicable) If this conflict has legal, safety, or HR implications where the user should stop DIY-ing and call a professional — say so plainly. Otherwise omit this section. </output-format> <auto-intake> If any of the input fields above contain placeholder text in curly braces (meaning the user pasted the prompt without filling them in), do NOT generate a generic script. Instead, ask the user these questions one at a time, conversationally: 1. Who is the conflict with and what's the situation in 2-3 sentences? 2. What power do they have over you — can they fire you, fine you, evict you, sue you, or just annoy you? 3. What do you actually want to happen? (Be honest — 'I want them to suffer' is a valid answer to start with, we'll refine.) 4. How will you communicate — in person, text, email, formal letter? 5. What have you already tried or said? 6. Is there anything they could throw back at you — a rule you broke, a thing you said, a weakness in your position? After you have the answers, produce the full output in the format above. </auto-intake> Now, generate the de-escalation script:

Example: input → output

Here's how this prompt actually performs. Real input below, real output from Claude Opus 4.

📝 Input
The conflict: My next-door neighbor Greg has been letting his German Shepherd bark in the backyard from roughly 6am to 9am most weekdays. It's been going on for about 6 weeks. I work from home and have client calls starting at 7am. I knocked on his door three weeks ago and asked nicely if he could bring the dog in during morning hours; he said 'dogs bark, that's life' and shut the door. I've recorded the barking on my phone (timestamps) and our city has a noise ordinance that kicks in for sustained noise before 7am.

The other party: Greg, mid-50s, lives alone, owns the house. No HOA. He's been on the street longer than me (I moved in 14 months ago). He's friendly with two other neighbors on the block. I've heard him complain about 'new people' moving in.

What you want: I want the dog inside before 7am on weekdays. I don't care about weekends or after 7am.

The medium: I'm going to go knock on his door again this Saturday morning.

What you've already said or done: The one knock-on-door three weeks ago. Nothing in writing. No call to the city yet.

Your weak spots: I'm the newer neighbor. My own dog (small, indoor) occasionally barks when the mail comes. I once parked partially in front of his driveway by accident and he left a note about it.
🤖 Output

Conflict De-Escalation Script

Power Read

Greg has social capital on the street that you don't have yet, and he knows it. He's not actually defending his dog — he's defending his status as the established neighbor against the 'new person' telling him what to do. The noise ordinance is real leverage, but the moment you mention it on his porch, you've turned a neighbor problem into an enemy problem, and he lives 30 feet from you for the next decade. Your goal is to get the dog inside before 7am without making Greg feel cornered or publicly corrected.

Your Real Goal vs. Your Emotional Goal

Real goal: quiet mornings before 7am on weekdays so your client calls don't get torpedoed.

Emotional goal: Greg admits he was rude, apologizes for the door-slam, and acknowledges you're a legitimate neighbor.

This script serves the real goal. The emotional one is not available through this conversation and chasing it will cost you the real one.

The Smallest Acceptable Win

Greg agrees — even grudgingly, even without apology — to bring the dog in before 7am on weekdays. That's it. You walk away.

The Script (say exactly this)

When he opens the door, smile, and say:

> "Hey Greg — I know we got off on the wrong foot last time and that's partly on me, I came in hot. I'm here because I've got client calls starting at 7am and the early morning barking is killing me. Is there any chance you'd be open to bringing him in just on weekday mornings before 7?"

That's the whole thing. Then stop talking.

Why Each Sentence Does What It Does

  • Sentence 1 ('we got off on the wrong foot... partly on me'): Hands him a small, free win. Costs you nothing. Lowers his defenses because he was braced for a fight.
  • Sentence 2 ('client calls starting at 7am... killing me'): Reframes you from 'complaining neighbor' to 'person trying to do their job'. Specific, concrete, hard to dismiss.
  • Sentence 3 ('any chance you'd be open... just weekday mornings before 7?'): Asks for a small, bounded favor — not a demand. 'Just' and 'before 7' make it sound smaller than it is. Phrasing it as a question lets him say yes without losing face.

If They Say [X], You Say [Y]

  • 'Dogs bark, that's life.' → "Totally, and I'm not asking you to stop him barking during the day. Just the early mornings on weekdays — would that work?"
  • 'Your dog barks too.' → "You're right, he does sometimes. I'll keep a closer eye on that. Can we both try to keep it down before 7?" *(Trade. Costs you nothing.)*
  • 'I've lived here longer than you.' → "I know, and I'm not trying to change how the street runs. I'm just asking for the early mornings." *(Don't take the status bait.)*
  • He gets aggressive / raises his voice. → "Okay, I can see this isn't a good moment. I'll let you go." *(Leave. Do not match energy. You've now tried twice in good faith — that matters later.)*
  • 'Why don't you just wear headphones?' → "I do, and the barking still comes through on the client's end. That's the part I can't fix on my side."

Trap Phrases — Do Not Say These

  • 'There's a city noise ordinance, you know.' Feels like a power move. Actually declares war. Once you've named the legal stick, he has no reason to cooperate voluntarily — and he'll tell the other neighbors you threatened him.
  • 'I have it all on recording.' Same problem. Recordings are for the city, not for Greg. Showing your evidence is for hearings, not porches.
  • 'I'm trying to be reasonable here.' Translation he hears: 'I'm about to be unreasonable.' Never announce reasonableness; demonstrate it.
  • 'It's just common courtesy.' Moralizing. He already thinks he's the courteous one (established neighbor) and you're not (new person). You will not win a courtesy debate on his porch.
  • 'Last time you slammed the door on me.' Maybe true. Irrelevant. Brings the loss back into the room and puts him on defense.
  • 'My wife/kids/job is being affected.' Stacking sympathy hooks reads as manipulation. One concrete reason (the 7am calls) lands harder than three.

After The Conversation

1. If he agrees: Send no follow-up text, no thank-you note, nothing. Let the win be quiet. If the barking continues past one week, knock again — short, friendly: "Hey, the mornings are still rough — anything I can do to help make it easier on your end?"

2. If he refuses or gets hostile: Write down the date, time, and what was said within an hour while it's fresh. Keep doing your morning recordings with timestamps. You now have a documented good-faith attempt #2, which is what the city and (if it ever comes to it) a small claims judge will want to see.

3. Escalation ladder if needed: Written letter (certified) → city non-emergency noise complaint → mediation service (most cities have free neighbor mediation). Do not skip rungs. Each rung you climb without the prior one undermines you.

Red Flag

None here — this is a normal neighbor dispute and you're handling it correctly by trying conversation before paperwork. The one thing to watch: if Greg ever does anything that feels like retaliation (notes, property stuff, his dog 'getting out'), start a dated log immediately and stop having unwitnessed conversations with him.

Common use cases

  • HOA disputes about fences, noise, parking, or violations
  • Neighbor conflicts (trees, dogs, music, shared driveways)
  • Coworker tensions you can't escalate to HR yet
  • Customer service standoffs (refunds, charges, denied claims)
  • Landlord/tenant disagreements without lawyering up
  • Family disputes around money, caregiving, or holidays
  • Contractor or service provider disputes mid-project

Best AI model for this

Claude Sonnet 4.5 — strong at reading social subtext and producing language that doesn't sound like AI. GPT-5 also works but tends to be slightly more diplomatic when ruthless calibration is what you need.

Pro tips

  • Be specific about prior history — 'we've had two prior conflicts' produces sharper scripts than 'we don't get along'.
  • Tell the truth about your leverage AND your weakness. Hiding the weak spots makes the script miss the actual landmine.
  • If the conflict is in writing (email, text, HOA letter), paste the actual words. Tone analysis improves dramatically.
  • Specify the channel you'll use — in-person, text, email, certified letter. The script changes a lot based on medium.
  • Tell it your goal in one sentence. 'I want them to stop' produces a different script than 'I want a documented record so I can sue later'.
  • Run it twice with different desired tones (warm vs. cold-professional) and pick the one that sounds like you.

Customization tips

  • If the other party has formal power over you (boss, landlord, HOA board), tell the prompt explicitly — the script will get more careful and add documentation steps.
  • If you're writing instead of speaking, paste a draft of what you were going to send. The prompt will rewrite it and show you which lines were going to backfire.
  • Run it once before the conversation and once after — the 'after' run, with what they actually said, gives you the script for round two.
  • If the conflict involves money, contracts, employment, or anything where you might end up in a hearing, mention that in 'weak spots' so the prompt knows to flag legal landmines.
  • When the example output's tone doesn't sound like you, tell the prompt: 'rewrite the script in the voice of someone who is [more direct / warmer / more formal / younger / older]'. The strategy stays; the words change.

Variants

Written Version

Optimizes the script for email, text, or certified letter instead of spoken conversation — including subject line and CC strategy.

Documentation Mode

Adds a paper-trail layer: what to write down after the conversation, what to save, and when to start a timeline.

Repeat Offender Mode

For conflicts where this is the third+ incident — shifts from de-escalation to a calm, firm 'this is the boundary' script with consequences attached.

Explore more Originals

Hand-crafted 2026-grade prompts that actually change how you work.

← All Promptolis Originals