mindfulness
À propos
Cette compétence fournit des techniques pour maintenir une conscience situationnelle et une clarté mentale dans des situations à haute pression ou potentiellement dangereuses. Elle inclut des cadres tels que le code couleur de Cooper et la boucle OODA pour l'évaluation des menaces, ainsi que des méthodes pratiques de désescalade et de gestion du stress. Les développeurs doivent l'utiliser lors de la programmation de scénarios impliquant la sécurité personnelle, l'analyse des risques environnementaux ou la résolution de conflits.
Installation rapide
Claude Code
Recommandénpx skills add pjt222/agent-almanac -a claude-code/plugin add https://github.com/pjt222/agent-almanacgit clone https://github.com/pjt222/agent-almanac.git ~/.claude/skills/mindfulnessCopiez et collez cette commande dans Claude Code pour installer cette compétence
Documentation
Cultivate Defensive Mindfulness
Develop applied situational awareness, de-escalation, mental clarity under threat — practical complement to seated meditation, operates in dynamic real-world env.
Use When
- Entering unfamiliar/potentially hostile env
- Assess social/physical situation for safety
- De-escalate verbal confrontation before physical
- Calm focus during high-pressure/dangerous event
- Rapid grounding after shock, surprise, adrenaline dump
- Daily movement awareness (walking, commuting, traveling)
- Mental prep before martial arts (
tai-chi,aikido)
In
- Required: Willingness for sustained outward attention (opposite of internal meditation)
- Required: Public/semi-public env access (streets, transit, events)
- Optional: Prior meditation (
meditate; helpful not required) - Optional: Martial arts background (
tai-chi,aikido; enhances physical response) - Optional: Practice partner for de-escalation role-play
Do
Step 1: Situational Awareness (Cooper Color Codes)
Calibrate awareness to environment.
Cooper Color Code Awareness Levels:
┌──────────┬─────────────────────┬──────────────────────────────────────────┐
│ Code │ State │ Description and Application │
├──────────┼─────────────────────┼──────────────────────────────────────────┤
│ White │ Unaware │ Absorbed in phone, headphones, day- │
│ │ │ dreaming. No awareness of surroundings. │
│ │ │ Acceptable only in secured private space │
├──────────┼─────────────────────┼──────────────────────────────────────────┤
│ Yellow │ Relaxed alert │ Aware of surroundings without fixation. │
│ │ │ Scanning people, exits, anomalies. This │
│ │ │ is the DEFAULT state in public spaces │
├──────────┼─────────────────────┼──────────────────────────────────────────┤
│ Orange │ Specific alert │ Something has triggered attention: a │
│ │ │ person, behavior, or situation. Forming │
│ │ │ a plan: "If X happens, I will do Y" │
├──────────┼─────────────────────┼──────────────────────────────────────────┤
│ Red │ Action │ The trigger condition from Orange has │
│ │ │ occurred. Execute the pre-formed plan. │
│ │ │ No hesitation — decision was made in │
│ │ │ Orange │
├──────────┼─────────────────────┼──────────────────────────────────────────┤
│ Black │ Overwhelmed │ Panic, freeze, tunnel vision. Caused by │
│ │ │ jumping from White directly to Red with │
│ │ │ no mental preparation. AVOID this state │
└──────────┴─────────────────────┴──────────────────────────────────────────┘
Practice protocol:
- Leaving home → consciously shift White → Yellow
- Yellow scan: who? exits? baseline behavior?
- Attention drawn → Orange: identify specific concern + form contingency
- Resolves → return Yellow. Don't stay Orange unnecessarily (drains energy)
- Practice White-to-Yellow transition until automatic (2-4 weeks daily)
→ Yellow = natural default in public. Anomalies register w/o conscious search. Exits + positioning habitual.
If err: Yellow exhausting/paranoid → too focused. Yellow = relaxed wide, peripheral not spotlight. Constantly Orange → over-calibrating threat. Practice safe familiar first → sustainable calm Yellow baseline.
Step 2: Body Language + Intent
Most threats broadcast intention before action. Learn pre-attack indicators.
- Baseline: new env → note normal: pace, posture, eye contact, group dynamics
- Deviation detection:
- Scanning crowd while still (target selection)
- Clenched fists, squared shoulders, bladed stance (pre-fight)
- Avoid eye contact + closing distance (predatory)
- Exaggerated calm/unnatural stillness in dynamic env
- Eyes: locked stranger eye contact = challenge or predatory focus. Repeated glance + away = surveillance/target assessment
- Proxemics: closing distance w/o social reason → attention. Trust "too close" instinct
- Group dynamics: one holds attention (distraction) + another maneuvers (setup). Pre-arranged signals (nods, gestures)
- Gut: limbic system processes threat faster than conscious. Feels wrong → honor signal, increase awareness before rationalizing
→ Notice pre-attack indicators real-time + Yellow → Orange w/ specific concern. Sense when behavior doesn't match social context.
If err: feels like guesswork → practice safe env first: cafe, transit, park. Note postures/distances/energy w/o threat. Built through observation volume. Hypervigilant (threats everywhere) → Step 6 grounding + recalibrate, most people aren't threats.
Step 3: Verbal De-escalation
Highest-value skill. Most violence preventable w/ words + positioning.
De-escalation Framework:
┌──────────────────┬──────────────────────────────────────────────────────┐
│ Phase │ Technique │
├──────────────────┼──────────────────────────────────────────────────────┤
│ 1. Space │ Maintain 2+ arm-lengths distance. Angle your body │
│ │ 45 degrees (non-confrontational, protects center │
│ │ line). Position an exit route behind you, never │
│ │ behind the aggressor │
├──────────────────┼──────────────────────────────────────────────────────┤
│ 2. Voice │ Lower your volume below theirs — this forces them │
│ │ to quiet down to hear you. Speak slowly. Use a │
│ │ calm, even tone. Avoid commands ("calm down") — │
│ │ use observations ("I can see you're upset") │
├──────────────────┼──────────────────────────────────────────────────────┤
│ 3. Acknowledge │ Name their emotion without agreeing with their │
│ │ position: "That sounds really frustrating." Do NOT │
│ │ say "I understand" unless you genuinely do. Do NOT │
│ │ argue, correct, or explain — yet │
├──────────────────┼──────────────────────────────────────────────────────┤
│ 4. Offer exits │ Give the person a way to disengage without losing │
│ │ face: "I think we both need a minute" or "Let me │
│ │ get someone who can help with this." Frame retreat │
│ │ as a mutual decision, not submission │
├──────────────────┼──────────────────────────────────────────────────────┤
│ 5. Disengage │ If de-escalation fails, create distance. Do not │
│ │ turn your back. Move toward other people, exits, │
│ │ or authority figures. Leave the area if possible │
└──────────────────┴──────────────────────────────────────────────────────┘
Critical:
- Never tell angry person "calm down" — escalates
- Hands visible + open (non-threatening) but positioned to protect (fence)
- Don't match energy: they escalate → you de-escalate harder
- Ego-driven need to "win" = most common cause of avoidable violence
→ Lower emotional temp via voice + positioning + verbal. Most de-escalate in 60-90s.
If err: physically threatening despite de-escalate → priority shifts to escape or physical defense (aikido, tai-chi). Not all situations talkable. Recognize failure + transition to action w/o hesitation.
Step 4: Moving Mindfulness
Apply meditation awareness to walking + commuting + public navigation.
- Walking → panoramic awareness: soften eyes, full visual field not single point
- Feel ground contact each step → anchors body while eyes scan
- Maintain awareness behind: sound changes (footsteps speeding, conversation stopping) carry info
- Transitions (entering building, rounding corner, stepping off transit) → brief pause + scan before commit
- Crowds → track 2-3 in periphery, no fixating
- "Mirroring walk": match pace + rhythm to blend in; vary pace to test if anyone matches
- Periodically: "If something happened now, where would I go?" = Yellow maintenance
→ Walking = active awareness, not passive transport. Transitions = natural scan points. Baseline maintained effortlessly.
If err: tiring/distracting → gripping too tight. Awareness = background music, present not demanding. Can't maintain w/ thinking/conversing → simple env first (quiet walk) before complexity (busy street, transit).
Step 5: Combat Mindfulness (OODA Loop)
OODA = Observe-Orient-Decide-Act. Decision cycle under pressure. Speed → who controls encounter.
OODA Loop Application:
┌──────────┬──────────────────────────────────────────────────────────────┐
│ Phase │ Application │
├──────────┼──────────────────────────────────────────────────────────────┤
│ Observe │ Take in the full situation: who, what, where, how many, │
│ │ weapons, exits, bystanders. Use peripheral vision. Do not │
│ │ fixate on the most obvious stimulus — scan the whole scene │
├──────────┼──────────────────────────────────────────────────────────────┤
│ Orient │ Match the observation to your training and experience: │
│ │ "This is [type of situation]. I have [these options]." │
│ │ Orientation is where pre-training pays off — trained │
│ │ responses orient faster than improvised ones │
├──────────┼──────────────────────────────────────────────────────────────┤
│ Decide │ Select the best available option — not the perfect one. │
│ │ A good decision now beats a perfect decision too late. │
│ │ If Orange-state planning was done (Step 1), the decision │
│ │ may already be made │
├──────────┼──────────────────────────────────────────────────────────────┤
│ Act │ Execute with full commitment. Hesitation between decision │
│ │ and action is the most dangerous gap. Once you act, observe │
│ │ the result and re-enter the loop │
└──────────┴──────────────────────────────────────────────────────────────┘
Training:
- Daily life → rapid scenario assessment: enter room, in 5s identify exits, cover, most concerning person
- "What if" games: "Aggressive entry that door now, what would I do?" Plan (Orange), release (Yellow)
- Martial arts → train pre-set responses to specific attacks → accelerates Orient
- Decision-making under artificial stress: timed drills, cold water + problem-solving, exercise + cognitive tasks
- Real/sim event → debrief: "Observed? Missed? Hesitated where?"
→ OODA increasingly automatic. Observation broad + rapid. Orient draws trained patterns. Decisions in Orange → Red action immediate.
If err: freeze under sim pressure (Black) → stimulus bypassed OODA. Gap White-to-Red too large. Step 1 + reinforce Yellow → unexpected meets alert mind. Freezing = normal survival, retrain w/ gradual stress inoculation, not extreme scenarios.
Step 6: Rapid Grounding
Stress/shock/adrenaline disrupts clarity → these restore awareness in seconds.
Grounding Techniques Quick Reference:
┌──────────────────────┬─────────────────────────────────────────────────┐
│ Technique │ Method and Use Case │
├──────────────────────┼─────────────────────────────────────────────────┤
│ Tactical breathing │ Inhale 4 counts, hold 4, exhale 4, hold 4. │
│ (box breathing) │ Repeat 4 cycles. Activates parasympathetic │
│ │ response in ~60 seconds. Use: acute stress, │
│ │ pre-confrontation, post-adrenaline dump │
├──────────────────────┼─────────────────────────────────────────────────┤
│ 5-4-3-2-1 sensory │ Name 5 things you see, 4 you hear, 3 you can │
│ anchor │ touch, 2 you smell, 1 you taste. Forces the │
│ │ mind out of internal panic and into present- │
│ │ moment external reality. Use: dissociation, │
│ │ freeze response, post-traumatic intrusion │
├──────────────────────┼─────────────────────────────────────────────────┤
│ Peripheral vision │ Fix eyes on a point, then widen awareness to │
│ activation │ the edges of the visual field without moving │
│ │ the eyes. Activates parasympathetic nervous │
│ │ system and reduces tunnel vision. Use: hyper- │
│ │ focus, tunnel vision, adrenaline narrowing │
├──────────────────────┼─────────────────────────────────────────────────┤
│ Physical anchor │ Press feet firmly into the ground and feel the │
│ │ contact. Squeeze and release fists 3 times. │
│ │ Roll shoulders back. These physical actions │
│ │ re-establish body awareness. Use: dissociation,│
│ │ shaking, post-event processing │
├──────────────────────┼─────────────────────────────────────────────────┤
│ Verbal reset │ State your name, location, and current task │
│ │ aloud: "I am [name], I am at [location], I am │
│ │ doing [task]." Orienting to facts breaks the │
│ │ emotional loop. Use: confusion, panic, sensory │
│ │ overload │
└──────────────────────┴─────────────────────────────────────────────────┘
- Practice NOT under stress → available when needed
- Tactical breathing = single most effective rapid intervention. Daily.
- Adrenaline event (near-miss, confrontation, shock) → box breathe before any decision
- Combine: physical anchor + tactical breathing for strong reactions
→ Acute stress → functional clarity in 60-120s. Practiced enough to recall under pressure w/o conscious effort.
If err: no relief in 2min → switch technique, not all work for all people/situations. Stressor ongoing (still in danger) → grounding premature, address situation via OODA (Step 5) first, ground after. Persistent inability post-event → professional support.
Step 7: Cross-Context Integration
Apply consistently across env.
- Urban: Higher baseline (solid Yellow). Track blind spots (alleys, stairwells, parking). Awareness at ATMs, transit platforms, vehicle entry/exit
- Wilderness: Different threat profile — terrain, weather, wildlife, getting lost. Awareness shifts people → environment. Navigation + shelter assessment replace social threat reading (bushcraft skills)
- Social events: Identify exits on arrival. Monitor others' alcohol (impaired = unpredictable). Closer to exits than crowd center
- Travel: Heightened in unfamiliar. Know emergency number for country. Distribute documents/valuables, not one bag. Note route to/from accommodation
- Digital: Awareness extends to info security — who's observing device, what shared publicly, physical security
- W/ others: Your awareness protects companions. Position between threats + companions. Brief on basic awareness w/o anxiety
→ Consistent sustainable baseline adapting to ctx, no paranoia/exhaustion. Yellow across env w/ Orange responses to genuine anomalies.
If err: anxiety/hypervigilance → calibration too high. Step 1, practice Yellow in familiar safe env. Goal = relaxed alertness, not perpetual scanning. Interferes w/ life enjoyment → mental health professional, esp. trauma history.
Step 8: Review + Refine
Skill improves through deliberate review + honest self-assessment.
- Notable awareness event (success, de-escalation, missed cue, freeze) → journal:
- What happened?
- Color code at start?
- Observed? Missed?
- Worked? Different next time?
- Monthly review: scan journal for patterns — recurring blind spots, env where awareness drops, emotional states that interfere
- Seek training: de-escalation workshops, scenario-based self-defense, first aid cert
- Partner practice: role-play confrontation, verbal de-escalation, critique positioning
- Cross-train: martial arts (
aikido,tai-chi) → physical response options; meditation (meditate) → calm baseline - Physical fitness: stress response performs better w/ conditioned cardiovascular
→ Measurable improvement over time: faster anomaly detection, calmer stress response, better positioning, more effective de-escalation.
If err: plateau → novel env or partners. Motivation wanes → awareness investment pays in one moment needed. Persistent weaknesses (always freeze, never notice rear approach) → target specifically not general practice.
Check
- Cooper codes identifiable + applicable real-time
- ≥3 pre-attack body language indicators named + recognized
- De-escalation framework articulable + practiced (min role-play)
- Moving mindfulness practiced daily commute/walking ≥1 week
- Box breathing performable from memory, daily practice
- ≥2 grounding techniques tested, one preferred identified
- OODA applied to ≥3 "what if" scenarios
- Awareness journal ≥3 real observations
Traps
- Hypervigilance as awareness: True awareness = relaxed sustainable. Exhausted/anxious/threats everywhere = chronic Orange — counterproductive + unsustainable. Yellow is goal, not Orange
- Tunnel vision on obvious threat: Yelling person = distraction. Train periphery scan when central attention grabbed. Multi-threat awareness = randori (
aikido) - "Calm down" to angry person: Most common de-escalation err. Communicates feelings invalid + you in control — both escalatory. Acknowledge emotional state instead
- Skip verbal before physical: Most violence preceded by verbal escalation. Effective de-escalation prevents most physical confrontations. Verbal skills > physical technique alone
- Skip post-event grounding: Adrenaline impairs judgment 20-45min post-event. Decisions (esp aggressive) unreliable in window. Ground first, decide second
- Train in isolation: Awareness + de-escalation = social skills. Solo builds foundation, partner drills + real-world = realistic competence
→
aikido— physical when de-escalation fails; blending + redirection mirrors verbaltai-chi— rooted calm + body awareness, supports physical readiness + emotional regulationmeditate— baseline mental stillness from which awareness operates; seated complements active outwardheal— first aid + stress management, direct applicationsremote-viewing— perceptual acuity training; non-local awareness complements environmental scanningawareness— AI self-application variant; Cooper codes + OODA → internal threat detection for hallucination + ctx degradation
Dépôt GitHub
Compétences associées
content-collections
MétaCette compétence propose une configuration éprouvée en production pour Content Collections, un outil axé sur TypeScript qui transforme des fichiers Markdown/MDX en collections de données typées de manière sûre avec une validation Zod. Utilisez-la lors de la création de blogs, de sites de documentation ou d'applications Vite + React riches en contenu pour garantir la sécurité de typage et la validation automatique du contenu. Elle couvre tout, de la configuration du plugin Vite et de la compilation MDX à l'optimisation des déploiements et la validation des schémas.
polymarket
MétaCette compétence permet aux développeurs de créer des applications avec la plateforme de marchés prédictifs Polymarket, incluant l'intégration d'API pour le trading et les données de marché. Elle fournit également une diffusion de données en temps réel via WebSocket pour surveiller les transactions en direct et l'activité du marché. Utilisez-la pour mettre en œuvre des stratégies de trading ou pour créer des outils traitant les mises à jour de marché en direct.
creating-opencode-plugins
MétaCette compétence aide les développeurs à créer des plugins OpenCode qui s'interconnectent avec plus de 25 types d'événements tels que les commandes, les fichiers et les opérations LSP. Elle fournit la structure du plugin, les spécifications de l'API événementielle et les modèles d'implémentation pour les modules JavaScript/TypeScript. Utilisez-la lorsque vous avez besoin d'intercepter, de surveiller ou d'étendre le cycle de vie de l'assistant IA OpenCode avec une logique personnalisée pilotée par les événements.
sglang
MétaSGLang est un framework de service LLM haute performance spécialisé dans la génération rapide et structurée pour les workflows JSON, regex et agentiques grâce à son cache de préfixe RadixAttention. Il offre une inférence nettement plus rapide, particulièrement pour les tâches avec des préfixes répétés, ce qui le rend idéal pour les sorties complexes et structurées ainsi que les conversations multi-tours. Choisissez SGLang plutôt que des alternatives comme vLLM lorsque vous avez besoin d'un décodage contraint ou que vous construisez des applications avec un partage étendu de préfixes.
