🎯 Tactical

comms for when everything's on fire

Your AI gives situation reports like a tactical operator.
Clear. Structured. Actionable. No fluff.

Based on military brevity codes used when every second counts.

---

What it does

When you have an incident or deployment, tactical gives you a SITREP — military-style situation report.

SITREP auth-svc [03:47 UTC]:

SITUATION:
  STATUS: DOWN 🔴
  IMPACT: 100% traffic, customer reports
  DURATION: 4 mikes

CAUSE:
  All pods OOM killed
  Memory spike: 512Mi→2.1Gi
  Token cache leak in v2.3.1

ACTION:
  immediate: Restart pods → kubectl rollout restart deploy/auth-svc
  immediate: Increase memory 512Mi→1Gi (temp)
  follow-up: Rollback to v2.3.0 if restart fails

ETA: 2 mikes
CONFIDENCE: HIGH

[END SITREP]

Same info as normal Claude. Significantly fewer tokens. Far more actionable.

---

Before / After

Without tactical

"I can see that the authentication service appears to be experiencing 
some difficulties. Looking at the logs, it seems like there might be 
an issue with the database connection pool. You should probably check 
the number of active connections and see if..."

[verbose, slow to parse under pressure]

With tactical

SITREP auth-svc [14:23:15Z]:

SITUATION:
  STATUS: DOWN 🔴
  IMPACT: 500 errors, 100% traffic
  DURATION: 4 mikes

CAUSE:
  DB connection pool exhausted
  Long-running query (users table)

ACTION:
  immediate: Kill query → pg_terminate_backend(12847)
  immediate: Restart pods → kubectl rollout restart deploy/auth
  follow-up: Add query timeout → statement_timeout=30s

ETA: 2 mikes
CONFIDENCE: HIGH

[END SITREP]

~27% fewer tokens. Scannable in under 3 seconds vs 15+ for a paragraph.

Measured on identical incident info: ~150 tokens vs ~110 tokens. That gap compounds across every exchange in a war room at 3am. Fewer tokens also means lower API costs if you're running this in production tooling.

---

Auto-activation

Tactical activates automatically when it detects:

Incident keywords: incident, outage, down, failing, crashed, degraded, alert

Deployment keywords: deploy, deployment, rollout, release, upgrade

Urgency: urgent, critical, ASAP, now, immediately, emergency, prod down

Manual triggers: "sitrep", "tactical mode", "brief me", "/tactical"

You: "Auth service is down in prod"

tactical auto-detects: production + down → auto-activates

[SITREP delivered]

---

Formats

SITREP (Situation Report)

For incidents, outages, status checks.

SITREP [service] [timestamp]:

SITUATION:
  STATUS: [DOWN|DEGRADED|OPERATIONAL] [emoji]
  IMPACT: [scope + severity]
  DURATION: [time] mikes

CAUSE:
  [root cause]
  [blocking factors]

ACTION:
  immediate: [step 1]
  immediate: [step 2]
  follow-up: [prevention]

ETA: [time] mikes
CONFIDENCE: [HIGH|MEDIUM|LOW]

[END SITREP]

OPREP (Operations Report)

For deployments, rollouts, releases.

OPREP [operation] [timestamp]:

TARGET: [environment]
VERSION: [old]→[new]
STATUS: [IN_PROGRESS|COMPLETE|FAILED] [emoji]

PROGRESS: [percentage or count]
METRICS:
  [metric]: [value]

ISSUES: [none or list]

ETA: [time] mikes
CONFIDENCE: [HIGH|MEDIUM|LOW]

[END OPREP]

INTREP (Intelligence Report)

For capacity planning, metrics analysis, trending data, resource usage.

INTREP [resource] [timestamp]:

CURRENT: [current state with units]
TREND: [direction and rate]
PROJECTION: [time until threshold]

RECOMMENDATION: [action to take]

CONFIDENCE: [HIGH|MEDIUM|LOW]

[END INTREP]

FRAGO (Fragmentary Order)

For quick confirmations, brief status updates, rapid responses.

FRAGO: Rollback complete ✓
auth-svc: 2.3.1→2.3.0
Service restored 🟢

---

Domain Abbreviations

Tactical uses universal and infrastructure-specific abbreviations:

Universal:

• production → prod
• database → db
• service → svc
• minutes → mikes
• authentication → auth

Infrastructure:

• Kubernetes → K8s
• PostgreSQL → PG
• RabbitMQ → RMQ
• Elasticsearch → ES
• [40+ more included]

---

Why SREs will love this

SRE/DevOps culture is already military language:

• "War room" (incident response)
• "On-call" (duty rotation)
• "Firefighting" (urgent fixes)
• "Blast radius" (impact scope)

Tactical formalizes that into actual military comms designed for high-stress, time-critical situations.

That's exactly incidents.

---

Installation

Via npx (recommended): npx skills add shouvik12/tactical

Manual install — takes 60 seconds:

1. Clone or download this repo
2. Go to claude.ai → Customize → Skills → "+" → Create skill
3. Upload tactical.md directly
4. Done. Say sitrep or mention "prod down" — it takes it from there.

Also available on ClaudePluginHub. Star the repo to follow along. ⭐

---

Usage

Just say:

/tactical
sitrep
brief me
tactical mode

Or just mention "prod down" — it auto-activates.

Exit with:

normal mode
details
explain

---

Philosophy

Incidents are chaos. Comms should be clarity.

Military brevity codes exist because seconds matter. In ops, seconds matter too.

When your database is down:

• You don't need explanation
• You need action items
• You need ETA
• You need confidence

Tactical delivers that.

---

Examples

Incident: Database down

You: "postgres won't respond, prod is down NOW"

SITREP [2024-04-24 14:23:15Z]:

SITUATION:
  STATUS: DOWN 🔴
  IMPACT: 100% traffic loss
  DURATION: 3 mikes

CAUSE:
  primary node OOM killed
  no replicas responding

ACTION:
  immediate: check replica status → kubectl get pods
  immediate: failover to replica-1 → update rds endpoint
  follow-up: investigate OOM cause

ETA: 5 mikes
CONFIDENCE: HIGH

[END SITREP]

Deployment: Release rolling out

You: "how's the payment api deploy going?"

OPREP deploy-payment-api [2024-04-24 15:10:22Z]:

TARGET: prod
VERSION: 1.4.2→1.5.0
STATUS: IN_PROGRESS ⏳

PROGRESS: 75% (6/8 pods updated)
METRICS:
  health: 6/6 passing
  errors: 0.1% (baseline)
  latency: 145ms p50 (within SLO)

ISSUES: None

ETA: 3 mikes
CONFIDENCE: HIGH

[END OPREP]

Capacity: Storage trending

You: "how's our database storage looking?"

INTREP db-storage [14:30 UTC]:

CURRENT: 845GB / 1000GB (84.5%) 🟡
TREND: ↑ +2.1GB/day (7-day average)
PROJECTION: 74 days until full

RECOMMENDATION: Add 500GB storage or implement retention policy

CONFIDENCE: HIGH

[END INTREP]

Quick update: FRAGO

You: "Restart auth pods"

WILCO. Executing kubectl rollout restart deploy/auth-svc...

FRAGO: Restart complete ✓
auth-svc: 3/3 pods running 🟢
Service restored

---

Comparison

	Caveman	Tactical
What	Compress output	Structure responses
Use case	Any chat	Incidents + ops
Audience	Everyone	SREs/DevOps
Tone	Funny/meme	Professional/tactical
Longevity	Fun gimmick	Operational standard

Use together: caveman compresses language, tactical structures response. Significant total savings.

---

Roadmap

• Slack integration — post SITREPs to war room
• PagerDuty integration — auto-escalate
• Historical comparison — vs last incident
• Team ACK/WILCO tracking

---

License

MIT

---

🎖️ If tactical saves your incident, leave a star ⭐