Agent Skills format - works with OpenClaw, Claude Code, Cursor, Codex, and other compatible clients | ClawHub

StrikeRadar by Yonatan Back tracks US-Iran tension across 8 data sources and aggregates them into a single risk score. This skill wraps that data so you can check it from the terminal or let your AI agent query it directly.

# ClawHub
clawdhub install strikeradar

# npx
npx skills add alexpolonsky/agent-skill-strikeradar

Or clone manually:

# OpenClaw
git clone https://github.com/alexpolonsky/agent-skill-strikeradar ~/.openclaw/skills/strikeradar

# Claude Code
git clone https://github.com/alexpolonsky/agent-skill-strikeradar ~/.claude/skills/strikeradar

# Cursor
git clone https://github.com/alexpolonsky/agent-skill-strikeradar ~/.cursor/skills/strikeradar

• Node.js 18+
• No dependencies

Ask your AI agent:

"What's the current Iran strike risk?"

Or run it as a standalone CLI:

npx tsx scripts/strikeradar.ts status

For quick access, add an alias to your shell profile (adjust the path to where you cloned it):

alias strikeradar='npx tsx ~/path/to/strikeradar/scripts/strikeradar.ts'

Then just:

strikeradar status
strikeradar signal news
strikeradar pulse

Timestamps are in UTC.

Valid signal names: news, connectivity, energy, flight, tanker, weather, polymarket, pentagon

  US Strike Radar  2026-02-20T07:35:47Z

  Total Risk: 32%
  ██████░░░░░░░░░░░░░░

  news            48%  23 articles, 11 critical
  connectivity    76%  CRITICAL (-61.6%)
  energy           0%  STABLE - $69.77 (+0.0%)
  flight           4%  87 aircraft, 8/8 key airlines
  tanker          10%  1 detected in region
  weather        100%  clear sky
  polymarket       4%  4% odds
  pentagon        10%  Low Activity

The API returns enough data for questions beyond "what's the overall risk score?":

• "Is something happening with Iran right now? Should I be worried?"
• "How many airlines are still flying over Iran?"
• "Is Iran's internet being shut down? Is it recovering or getting worse?"
• "How many people are watching the strike radar right now?" (a surge usually means something is developing)
• "What are the latest headlines about a US strike on Iran?"
• "Is the situation escalating or calming down?"
• "Are there military refueling tankers near Iran right now?"
• "What do Polymarket bettors think about a US strike on Iran?"
• "Is there unusual late-night activity near the Pentagon?" (the famous Pizza Index)
• "Is the oil price spiking right now?"

With agents that support scheduled tasks (e.g., OpenClaw cron), you can set up ongoing monitoring:

• "Alert me if the Iran strike risk crosses 50%"
• "Send me a morning briefing every day with all signals and what changed overnight"
• "Warn me at 50%, 65%, and 80% risk levels"
• "Let me know when the situation calms down after a spike"
• "Notify me if Iran's internet starts going down" (often a precursor to escalation)
• "Tell me if there's unusual late-night Pentagon activity"

Calls the public API at api.usstrikeradar.com and formats the response. No scraping, no auth, no API key needed. Outputs human-readable text in the terminal and JSON with next_actions when piped.

• Data refreshes roughly every 30 minutes on the server side
• Risk scores are algorithmic estimates, not intelligence assessments
• No historical data beyond the rolling history array per signal
• Viewer pulse data is approximate

Independent open-source tool wrapping the StrikeRadar API by Yonatan Back. Not affiliated with or endorsed by StrikeRadar. Probability scores are algorithmic estimates from publicly available data, not intelligence assessments. Do not use for personal, financial, or safety decisions. Provided "as is" without warranty of any kind.

Alex Polonsky - GitHub - Twitter/X - LinkedIn

Part of Agent Skills - Specification