Unphurl

# Unphurl > Domain intelligence API for AI agents and developers One URL in. Structured signals out. Your AI agent decides what to do with them. ## What it is Unphurl analyses URLs across 7 dimensions and returns structured, scored signals. Not binary verdicts. Built for autonomous AI agents, security bots, and developers who need to understand what's behind any URL before acting on it. ## Key capabilities - 7 analysis dimensions: redirects, brand impersonation, domain intelligence (RDAP), SSL/TLS, parked detection, URL structure, DNS enrichment - Brand impersonation detection against 150+ brands (Levenshtein + homoglyph analysis) - 23 configurable scoring weights with custom profiles per use case - Score breakdown transparency (see exactly which signals contributed) - Batch processing up to 500 URLs with async polling and webhooks - Usage analytics per account (free rate, pipeline spend, score distribution) ## Three interfaces - REST API: https://api.unphurl.com/v1 (bearer auth) - CLI: npx unphurl (colour-coded output, batch processing, account management) - MCP server: @unphurl/mcp-server (13 tools for Claude Code, Cursor, Windsurf) ## Claude Cowork Connector URL: https://mcp.unphurl.com/mcp Add via: Settings > Connectors > Customize > Add Custom Connector ## Pricing No subscriptions. Credits never expire. Most lookups are free. - Starter: $9 for 100 pipeline checks ($0.090/check) - Standard: $39 for 500 pipeline checks ($0.078/check) - Pro: $99 for 2,000 pipeline checks ($0.050/check) - Scale: $399 for 10,000 pipeline checks ($0.040/check) 20 free pipeline checks on signup. No credit card required. ## Three-gate billing 1. Known domain (Tranco Top 100K)? Free, instant, score 0. 2. Previously analysed (cached)? Free, instant, scored with your weights. 3. New domain? Full 7-dimension pipeline. 1 credit. Results cached for all future lookups. Typical free rate exceeds 90%. ## How it differs from alternatives Unlike Google Safe Browsing and VirusTo