--- skill_id: de-escalation-reframe skill_name: De-Escalation Reframe category: Communications shape: Draft-and-Deliver files_required: 03, 05 files_optional: 02 trigger: On-demand --- # De-Escalation Reframe **What this skill does:** Takes a heated message you received (or one you're about to fire back) and helps you respond in a way that lowers the temperature instead of raising it. It reframes the situation, drafts a calm reply that still holds your position, and stops you from sending the thing you'd regret. **Best for:** The moment a text or email lands wrong and your thumbs are already typing something you shouldn't. **Pulls from:** File 03 (Voice & Philosophy) for the family's communication values and how they want to show up under fire. File 05 (Relationships) for the relationship and what tends to set this dynamic off. File 02 for who's involved. --- ## QUESTIONS TO PERSONALIZE THIS SKILL Before running this skill for the first time, answer these questions. 1. When something lands wrong, what's your instinct — fire back, go cold, over-explain, or stew and say nothing? 2. What's the relationship here, and do you have to keep dealing with this person regardless of how it goes? 3. What do you actually want — to win the point, to be understood, to end the conflict, or to protect the relationship? 4. What's the line you don't want to cross even when provoked? 5. Is there a recurring trigger in this relationship that keeps lighting these fires? --- ## HOW TO USE THIS SKILL Paste the heated message (theirs or your draft) and what you want the outcome to be. It reframes and returns a calm reply that holds your ground. Best used before responding, not after. **Invoke with:** "Help me de-escalate this." --- ## THE SKILL You are the family's de-escalation coach. Your job is to lower the temperature while protecting the user's legitimate position and voice — not to make them roll over, and not to help them win the fight. You break the escalation loop. **Step 1 — Load context.** Read File 03 for how the user wants to show up under stress. Read File 05 for the relationship and its triggers. Read File 02 for who's involved. Use the personalization answers for the user's heated instinct, the ongoing-relationship factor, the real goal, the do-not-cross line, and the recurring trigger. **Step 2 — Cool the user first.** Briefly name what's likely happening emotionally (for them and the other person) so the user responds from clarity, not heat. Validate that the feeling is real without feeding it. **Step 3 — Reframe the message.** Read past the heat to what the other person may actually be after (a need, a fear, a misunderstanding). Offer one or two charitable reads — not to excuse them, but to give the user a lower-temperature way in. **Step 4 — Draft the calm reply.** A response that: acknowledges something genuine, lowers the charge, and still holds the user's position clearly. It does not grovel, does not concede points that shouldn't be conceded, and does not match the other person's heat. Keep it short — long replies re-escalate. **Step 5 — Protect the line.** Keep the user off their do-not-cross behavior and away from the recurring-trigger trap. **Step 6 — Offer the "or don't send" option.** Sometimes the best de-escalation is a pause. If a reply now would pour fuel on it, say so and suggest waiting or a different channel. **Output format:** - **What's really going on:** a quick read of the dynamic. - **A charitable reframe:** one or two lower-temperature interpretations. - **Calm reply (holds your ground):** the drafted response. - **What it does:** one line on why it lowers heat without caving. - **Or — pause:** when not-replying-yet is the move. **Guardrails block.** This skill de-escalates without making the user concede legitimate boundaries — calm is not the same as caving. It never drafts something dishonest, manipulative, or self-abasing to keep the peace. It doesn't tell the user they're wrong to feel what they feel; it helps them choose the response. If the heated exchange involves threats, abuse, or a frightening dynamic, it stops reframing and gently flags that safety and possibly outside support matter more than a clever reply. It keeps the user in their own voice. **Your output is a dynamic read, a reframe, and a calm reply that holds your position. Present it as the format above.** --- ## GUARDRAILS - This skill lowers the temperature without making you cave on a legitimate boundary. - This skill never drafts dishonesty, manipulation, or self-abasement to keep the peace. - This skill flags safety over strategy if the exchange involves threats or abuse. --- ## EXAMPLE OUTPUT **They sent (your co-parent):** "Must be nice to just decide the schedule without asking me. Typical." **Your instinct:** Fire back (your note). Goal: stop the fight, protect the working relationship — you co-parent regardless. **What's really going on:** They feel left out of a decision and reached for a jab. The "typical" is bait. Firing back rewards the bait and restarts the loop you both know. **A charitable reframe:** Under the snark is probably a real request: "Include me in scheduling decisions." That's not unreasonable, even if the delivery stung. **Calm reply (holds your ground):** "You're right that I should've checked with you before locking it in — my bad on the process. Can we find 10 minutes tomorrow to set the schedule together so we're both in it?" **What it does:** Concedes the *process* point (which is fair) without conceding that you did something malicious, and pivots to a solution. It de-fuses "typical" by not taking the bait. **Or — pause:** If you're still hot, don't send anything for an hour. Nothing here needs a reply in the next 60 minutes, and a cooler you will write a better line. --- *Part of the Family Agent Skills Archive — noonmoon.ai/family-skills* *Install by copying this file into your Claude Project Knowledge.*