Atlas / Skills / leoyeai / Physical De Escalation

Physical De EscalationSAFE

skills/leoyeai/physical-de-escalation

🧠 Curated collection of 1209+ best OpenClaw skills — weekly updated by MyClaw.ai

Verdict
SAFE
Grade
B
Trust score
89 /100
Version
—
Hosts
1 documented
License
MIT
Stars
2,160
01

Overview

🧠 Curated collection of 1209+ best OpenClaw skills — weekly updated by MyClaw.ai

Read from source at commit 4f3b4a2a472eOBSERVED · 2026-10-08
02

Host compatibility

What the documentation claims. We have not run a compatibility test.

HostStatusNotes
openclawmentioned
03

What it tells the agent

The instruction file, verbatim from the audited commit — this is the text the model reads, and the surface the audit's instruction layer examines. Quoted here so you can judge it without cloning anything.

---
name: physical-de-escalation
description: >-
  De-escalation techniques for in-person confrontations and tense physical situations. Use when someone faces angry customers, aggressive strangers, workplace confrontations, or needs to calm a heated situation in a physical space.
metadata:
  category: skills
  tagline: >-
    Defuse face-to-face confrontations — angry customers, aggressive strangers, heated arguments — using body language, tone, and positioning.
  display_name: "Physical Space De-escalation"
  submitted_by: HowToUseHumans
  last_reviewed: "2026-03-19"
  openclaw:
    requires:
      tools: [filesystem]
    install: "npx clawhub install physical-de-escalation"
---

# Physical Space De-escalation

Most conflict advice focuses on what to say. This skill is about what to do with your body when you're in a room with someone who is angry, aggressive, or escalating. Body language, positioning, voice, and physical distance do more to de-escalate a confrontation than the perfect script. Retail workers, bartenders, nurses, teachers, and parents use these techniques every day. They work because they address the physiology of aggression — a person in fight-or-flight mode is responding to physical cues before they process words. Get the body language right and the words matter less. Get it wrong and even perfect words won't help.

```agent-adaptation
# Localization note — cultural norms around conflict vary significantly
- Personal space expectations differ:
  US/Northern Europe: ~4 feet conversational distance
  Latin America/Middle East/Southern Europe: ~2-3 feet
  East Asia: varies, but physical contact between strangers is rare
- Eye contact norms:
  US/Western Europe: direct eye contact signals confidence/honesty
  East Asia/some Indigenous cultures: prolonged eye contact can
  signal aggression or disrespect
  Adapt eye contact advice to cultural context.
- Physical gestures: open palms are near-universal as non-threatening,
  but specific gestures vary (thumbs-up offensive in some cultures, etc.)
- Legal self-defense frameworks vary by jurisdiction:
  US: varies by state (stand your ground vs. duty to retreat)
  UK: "reasonable force" standard
  Many jurisdictions: duty to retreat if safe to do so
- Emergency numbers: US 911, UK 999, AU 000, EU 112
```

## Sources & Verification

- **Crisis Prevention Institute (CPI)** -- Verbal and nonverbal de-escalation model used in healthcare, education, and social services. https://www.crisisprevention.com/
- **Gavin de Becker, "The Gift of Fear"** -- Research on pre-violence indicators and trusting survival instincts. Published 1997, remains the standard reference.
- **OSHA** -- Workplace violence prevention guidelines. https://www.osha.gov/workplace-violence
- **Law enforcement verbal judo / tactical communication** -- George Thompson, "Verbal Judo: The Gentle Art of Persuasion." Core principles adapted for civilian use.
- **Anthropic, "Labor market impacts of AI"** -- March 2026 research showing this occupation/skill area has near-zero AI exposure. https://www.anthropic.com/research/labor-market-impacts

## When to Use

- User is a retail or food service worker dealing with angry customers
- User is in a confrontation with an aggressive stranger
- User needs to calm a heated argument at a family gathering
- User is a bartender, bouncer, or security worker
- User faces a road rage situation
- User is a nurse, teacher, or social worker dealing with agitated individuals
- User witnessed a confrontation and wants to know how to intervene safely
- User wants to prepare for situations that could turn physical

## Instructions

### Step 1: The 5-second assessment

**Agent action**: Teach the user how to rapidly read a situation before responding.

```
5-SECOND THREAT ASSESSMENT:

When you encounter an agitated person, scan these five things
before you say or do anything:

1. HANDS — What are they doing with their hands?
   Open and visible = lower threat
   Clenched fists = escalating
   Hands hidden / reaching into pockets or waistband = high alert
   Holding an object (bottle, tool, bag) = potential weapon

2. STANCE — How is their body positioned?
   Squared up facing you directly = confrontational
   Bladed (one foot forward, angled) = preparing for action
   Pacing or bouncing = adrenaline surge, escalating
   Seated or leaning back = lower threat

3. FACE — What does their expression tell you?
   Jaw clenched, nostrils flared = anger, escalating
   Thousand-yard stare / flat affect = potentially most dangerous
   Eyes scanning for witnesses/exits = calculating
   Crying or trembling = distress, not necessarily aggression

4. VOICE — What's happening with their speech?
   Loud and fast = adrenaline, fear, anger
   Quiet and controlled + tense body = potentially more dangerous
     than someone yelling
   Repeating the same phrase = stuck in a loop, not processing
   Slurred or erratic = substance influence (changes approach)

5. CONTEXT — What's the environment?
   Are there exits? Other people? Objects that could be weapons?
   Is the person cornered? (Cornered people are more dangerous.)
   Are there children or vulnerable people present?
   Is this person known to you, or a stranger?

THREAT LEVEL DECISION:
LOW: Person is upset but in control. Approach and de-escalate.
MEDIUM: Person is escalating, aggressive posture, yelling.
  De-escalate but maintain distance and plan your exit.
HIGH: Pre-attack indicators present (see below). Do not engage.
  Create distance. Call for help. Leave if possible.
```

### Step 2: Body positioning and distance

**Agent action**: Teach physical positioning for de-escalation.

```
YOUR BODY POSITION — THE MOST IMPORTANT VARIABLE:

THE 45-DEGREE ANGLE:
- NEVER stand face-to-face with an agitated person. Squaring up
  is a primate dominance signal. It triggers escalation.
- Stand at a 45-degree angle to them. One foot slightly forward,
  body angled. This is non-confrontational and also gives you
  better
04

Trust audit

SAFEgrade B · trust 89/100 Nothing in the source contradicts what it says it does. Grade A is reserved for packages that have also passed the behavioural sandbox.

LayerWhat it checksResult
L0Provenance & inventoryPASS
L1Static analysis of the codeWARN
L2Instruction surface (what it tells the agent)PASS
L3Class-specific surfacePASS
L4Behavioural (sandbox)SKIPPED

What the source does

Filesystem
none-observed
Network
none-observed
Shell
none-observed
Dependencies
pinned
Secrets in source
none-found

Findings (1)

MEDIUMObfuscation / stealth · obf.base64_blob · CWE-506, CWE-94
skills/compdf-conversion-cli/scripts/license.xml:9
<key>k5Ey9KFlkqpj+SDkUw+5ED9lTA3En/qUi0zdrydUCH3kMWTE3Eh65NXnFCaxlY2omY2JHnlEoK7Li7oOEvM7eG5VPdcO/sFlMfoCRdnLYdepJ+uLzYwOWR8W4yQVve/clxVFTVRL4DFleKInGdpAxIbHZT2yi4ADAMENls1N1XSLojRuqXePXDeAT/4Mv4TTx0s

Gates applied: no_behavioural_pass.

Audited 2026-10-08 · audit v0.4.1 · source sha 4f3b4a2a472efull audit observations/trust-audit/skill/leoyeai__physical-de-escalation.json · Report an issue / request a re-scan
05

Audit history

Every audit this skill has had.

DateSourceVerdictGradeScoreChange
2026-10-084f3b4a2a472eSAFEB89first audit
06

Questions

What does the Physical De Escalation skill do?

🧠 Curated collection of 1209+ best OpenClaw skills — weekly updated by MyClaw.ai

Is Physical De Escalation safe to install?

The audit found nothing in the source that contradicts what it says it does, and graded it B (89/100). Grade A is held back for packages that have also passed a sandboxed behavioural run, which is why a clean skill reads B.

What can Physical De Escalation access on my machine?

The audit observed no filesystem, network or shell use at all in its source.

Which assistants does Physical De Escalation work with?

Its documentation mentions openclaw. That is what the text claims, not a compatibility test we ran.

How current is this page?

The grade is for one exact copy of the source (4f3b4a2a472e), read on 2026-10-08. The repository is watched, and a new audit runs when it changes — this is the first audit.

Advertisement