A live intelligence feed for the Iran / Middle East conflict. Built to answer one question: what is happening right now, and what do I need to do?
KONFLIKT scans live news sources every 5 minutes using the Claude AI API with web search, then builds a structured, human-readable intelligence briefing — automatically. The goal is to give anyone, regardless of background knowledge, a complete picture of what's happening in under 60 seconds of reading.
It was designed as a life-safety awareness tool: if you're in or near a conflict zone, or you have family there, this tells you what's happening, what's at risk, where to go, and what comes next.
The first thing you see after a scan. Designed for maximum speed of comprehension:
- 3 breaking facts — the most urgent things happening right now, no fluff
- Immediate Danger — one sentence on the biggest physical threat at this moment
- Recommended Action — one sentence on what to do or avoid
- Full Situation Narrative — 4 paragraphs written for someone with zero background: what caused this, what happened in the last 48 hours, what's happening right now, and what civilians need to know
- Top Prediction — the single most likely next event with probability percentage
A live event log of everything that has happened, ranked by severity (CRITICAL → LOW). Each event is expandable with a full 2-3 sentence explanation. Categorized by type: STRIKE, DIPLOMATIC, MOVEMENT, STATEMENT, ECONOMIC.
Who is on whose side — updated with every scan.
- Side-by-side scoreboard: US Coalition vs Iran Axis vs Neutral
- Every country/group tracked individually with:
- Current alignment (US_COALITION / IRAN_AXIS / NEUTRAL / SHIFTING)
- Their specific role (Active combatant, Arms supplier, Airspace access, Proxy force, Mediator, etc.)
- Status: ACTIVE, SUPPORTING, WATCHING, or SHIFTING
- One-sentence note on what they are doing right now
- Purple alert banner when any country's alliance is in flux
- Filter by side to focus on one coalition
Every location in the conflict zone rated:
- 🔴 DANGER — active strikes, fighting, or direct threat
- 🟠 CAUTION — elevated risk, possible target
- 🟡 WATCH — monitoring, situation developing
- 🟢 SAFE — currently stable
Sorted by threat level, with a plain-language reason for each rating.
What is at risk right now and how likely it is:
- Targets: infrastructure, personnel, economic, diplomatic
- Likelihood: HIGH / MEDIUM / LOW
- Sorted by urgency
A live probability model of what comes next. Each scenario shows:
- Probability percentage (0–100%)
- Timeframe (24h / 48h / 1 week)
- Trend arrow (▲ increasing / ▼ decreasing / ● stable)
- Reasoning based on current events
Probabilities shift with every scan as new intelligence comes in.
The complete history of how we got here — from 1953 to today.
29 fully written historical entries covering every major turning point:
- 1953 CIA coup that started the US-Iran rift
- 1979 Islamic Revolution and hostage crisis
- 2003 Iraq invasion that accidentally empowered Iran
- 2010 Stuxnet cyberattack
- 2015 nuclear deal → 2018 Trump withdrawal
- 2020 Soleimani assassination
- 2024 first-ever direct Iran → Israel missile attacks
- 2024 Nasrallah killed, Assad regime collapses
- 2025–2026 current war escalation
Filter by category: ☢ Nuclear · ⚔ Military · 🏛 Political · 💰 Economic · 🕸 Proxy War
Each entry is expandable with a full explanation of what happened and why it matters — written so anyone can understand the chain of cause and effect. Timeline ends with a pulsing "YOU ARE HERE" marker.
- Historical background paragraph
- Iranian civilian sentiment vs. the regime
- Key players directory (who they are and what role they play)
Once initialized, the app automatically re-scans every 5 minutes. A green countdown bar in the header shows time until the next refresh. You can also hit REFRESH NOW at any time to force an immediate update.
Every refresh updates:
- Threat level
- Breaking facts
- Full narrative
- Events log
- Alliance map
- Zone statuses
- Threat assessments
- Prediction probabilities
| Layer | Technology |
|---|---|
| Frontend | React (single .jsx file) |
| AI / Web Search | Anthropic Claude API (claude-sonnet-4-20250514) with web_search_20250305 tool |
| Styling | Inline CSS + Google Fonts (Share Tech Mono, Oswald) |
| Hosting | Runs as a Claude.ai artifact, or any React environment |
| Data | 100% live — no database, no backend, no static data (except the historical timeline) |
User clicks "Initialize Scan"
↓
POST to Anthropic /v1/messages
- Model: claude-sonnet-4-20250514
- Tool: web_search (searches live news)
- System prompt: structured JSON schema
↓
Claude searches: "Iran military news", "Israel Iran conflict",
"Middle East war", "US military Middle East",
"Hezbollah Houthis latest"
↓
Returns single JSON object with all sections
↓
React renders all 7 tabs from that one response
↓
Timer fires every 5 minutes → repeat
- Open claude.ai
- Paste the full
.jsxfile as a prompt asking to render it as a React artifact - Click Initialize Scan
npx create-react-app konflikt
cd konflikt
# Replace src/App.js with the contents of conflict-tracker.jsx
npm startNote: The Anthropic API key is handled automatically inside Claude.ai artifacts. For standalone deployment you will need to add your own API key to the fetch headers and set up a backend proxy (never expose API keys in frontend code).
conflict-tracker.jsx # Entire app — single file
README.md # This file
The entire application is intentionally a single file. No build step required inside Claude.ai. For production deployment, splitting into components is recommended.
If you want to rebuild this from scratch using Claude, use this prompt:
Build a single-file React JSX app called KONFLIKT — a real-time conflict intelligence
dashboard for the Iran/Middle East war.
The app calls the Anthropic API at https://api.anthropic.com/v1/messages using
model claude-sonnet-4-20250514 with the web_search_20250305 tool enabled.
No streaming — standard POST fetch only. No API key in headers (handled externally).
The API call returns a single JSON object with these fields:
threat_level, headline, breaking (array of 3), immediate_danger, safe_direction,
narrative (4 paragraphs), events (array), zones (array), threats (array),
predictions (array with probability 0-100 and trend), background,
iranian_sentiment, key_players (array), alliances (array with side/role/status/note).
The app has 7 tabs:
1. TIMELINE — hardcoded historical events 1953–2026, filterable by category
(nuclear/military/political/economic/proxy), vertical timeline layout, expandable cards
2. LIVE BRIEF — breaking facts, danger/safe boxes, 4-paragraph narrative, top prediction
3. EVENTS — expandable event log sorted by severity with category icons
4. ALLIANCES — US COALITION vs IRAN AXIS scoreboard, country cards showing
side/role/status/note, SHIFTING alert banner, filterable by side
5. ZONES — grid of locations rated DANGER/CAUTION/WATCH/SAFE
6. THREATS — cards sorted by HIGH/MEDIUM/LOW likelihood
7. PREDICTIONS — probability bars with trend arrows, sorted by probability
Auto-refreshes every 5 minutes with a countdown bar in the header.
Shows threat level badge, breaking headline ticker, and last-updated timestamp.
Design: dark war-room aesthetic, #080808 background, Share Tech Mono + Oswald fonts,
color coding: red=#ff3300 (critical/Iran), blue=#3399ff (US coalition),
green=#00dd66 (safe/low), white text for all headings and primary content.
This tool is for awareness and safety. It is designed to:
- Help people near conflict zones understand the situation quickly
- Track which areas are dangerous and which are safer
- Show where alliances stand so you know which countries are safe to travel to or through
- Give a prediction of what comes next so you can prepare
It is not a classified intelligence tool. All data comes from open-source news via web search. Treat predictions as informed estimates, not certainties.
This application uses AI-generated analysis of open-source news. It is not affiliated with any government, military, or intelligence agency. Predictions and threat assessments are probabilistic estimates based on publicly available information — not guarantees. Always consult official government travel advisories and emergency services in life-threatening situations.
MIT — use freely, modify freely, deploy freely.
Built with Claude by Anthropic · KONFLIKT Intel Feed · Open Source Intelligence · Not Classified