JSPM

  • ESM via JSPM
  • ES Module Entrypoint
  • Export Map
  • Keywords
  • License
  • Repository URL
  • TypeScript Types
  • README
  • Created
  • Published
  • Downloads 3
  • Score
    100M100P100Q46902F
  • License MIT

Production monitoring built for developers who ship fast. Auto-diagnoses issues with Claude AI.

Package Exports

  • scanwarp
  • scanwarp/dist/index.js

This package does not declare an exports field, so the exports above have been automatically detected and optimized by JSPM instead. If any package subpath is missing, it is recommended to post an issue to the original package (scanwarp) to support the "exports" field. If that is not possible, create a JSPM override to customize the exports field for this package.

Readme

ScanWarp CLI

Your AI writes your code. ScanWarp keeps it running.

Production monitoring built for developers who ship fast with AI tools like Cursor and Claude Code. Auto-diagnoses issues and suggests fixes directly to your AI coding assistant.

License: MIT npm version

Quick Start

# Monitor locally while you build
npx scanwarp dev

# Connect production app to ScanWarp
npx scanwarp init

What You Get

After running scanwarp init, you'll have:

📊 Web Dashboard

Visit your server URL (e.g., http://localhost:3000 or your hosted URL) to access:

  • Overview - System health at a glance
  • Monitors - Uptime and response time tracking
  • Events - Real-time feed from your app, Stripe, GitHub, Vercel
  • Incidents - Auto-detected issues with AI diagnosis
  • Traces - OpenTelemetry request waterfalls with bottleneck highlighting

🤖 AI Diagnosis

When something breaks, Claude AI analyzes the full context and tells you:

  • Root cause in plain English
  • Exact file and line that needs fixing
  • Ready-to-execute fix prompt for your AI tool

🔌 MCP Integration

Your AI coding tool (Cursor, Claude Code) connects via MCP to:

  • See what's broken
  • Get fix suggestions
  • Read trace details
  • Resolve incidents

All without leaving your editor.

📡 Request Tracing

Zero-config OpenTelemetry instrumentation captures:

  • HTTP requests
  • Database queries
  • External API calls
  • Performance bottlenecks

Automatically installed with scanwarp init (or skip with --skip-instrumentation).

🔔 Notifications

Get alerts via:

  • Discord webhooks
  • Slack webhooks
  • Email (coming soon)

Commands

Command Description
scanwarp dev Run full monitoring flywheel locally while you build
scanwarp init Connect production app to ScanWarp server
scanwarp server Self-host the server with SQLite (zero deps)
scanwarp status Check monitor health and active incidents
scanwarp events Stream events (--follow, --type, --source)

What Gets Monitored

Source What it captures
Health checks HTTP monitoring every 60s, response time
Vercel Production errors via log drain
Stripe Payment failures, subscription issues
GitHub Failed CI, Dependabot alerts
Supabase Database health, connection pool
OpenTelemetry Request traces (HTTP, DB, Redis)
Provider status Vercel, Stripe, GitHub, Cloudflare, Railway, AWS outages

During scanwarp dev, also detects N+1 queries, slow queries, and schema drift.

Deploy Your ScanWarp Server

You need a ScanWarp server to send data to. Choose one:

Option 1: One-Click Deploy (60 seconds)

Deploy on Railway Deploy to Render

Then run: npx scanwarp init --server https://your-server-url.up.railway.app

Option 2: Self-Host

# SQLite, zero dependencies
npx scanwarp server

# Or with Docker + Postgres
docker compose up -d

Then run: npx scanwarp init --server http://localhost:3000

Connect Your AI Tool

Cursor

  1. Open Cursor Settings → Features → MCP
  2. Add ScanWarp MCP server:
{
  "mcpServers": {
    "scanwarp": {
      "command": "npx",
      "args": ["-y", "scanwarp", "mcp", "--server", "YOUR_SERVER_URL"]
    }
  }
}
  1. Restart Cursor
  2. Ask: "What's broken in production?" or "Show me the slowest requests"

Claude Code

Run scanwarp init and select "Yes" when asked about MCP configuration. It will auto-configure for you.

Environment Variables

Variable Description
SCANWARP_SERVER_URL Your ScanWarp server URL
SCANWARP_PROJECT_ID Project identifier (auto-saved after init)
ANTHROPIC_API_KEY Required for AI diagnosis

Example Workflow

# 1. Deploy ScanWarp server
# (Use Railway/Render button above)

# 2. Initialize monitoring in your app
cd my-nextjs-app
npx scanwarp init --server https://my-scanwarp.up.railway.app

# 3. Deploy your app
git push

# 4. Check the dashboard
# Visit https://my-scanwarp.up.railway.app

# 5. Ask your AI tool (via MCP)
# "What's broken in production?"
# "Show me the slowest API endpoints"
# "Get the fix prompt for incident #123"

Learn More

License

MIT