@unphurl/mcp-server

@unphurl/mcp-server

Domain intelligence for AI tools. Unphurl analyses URLs across seven dimensions (redirect behaviour, brand impersonation, domain intelligence via RDAP, SSL/TLS validity, parked domain detection, URL structural analysis, and DNS enrichment) and returns structured signals with a configurable risk score. 23 scoring weights, all customizable per use case. One URL in, structured intelligence out. Your agent decides what to do with it.

What it costs

Every new account gets 3 free pipeline check credits to test with real URLs. Most lookups are free. Known domains (Tranco Top 100K) and previously analysed domains return cached results at no cost. You only pay when an unknown domain runs through the full analysis pipeline. Packages start at $9 for 100 pipeline checks (one-time purchase, no subscription). In typical use, 95-99% of URLs resolve free.

Quick start

1. Add to your MCP configuration

Add to your .mcp.json (Claude Code, Claude Desktop, Claude Cowork, ChatGPT desktop, Cursor, Windsurf, or any MCP-compatible tool):

{
  "mcpServers": {
    "unphurl": {
      "command": "npx",
      "args": ["-y", "@unphurl/mcp-server"],
      "env": {
        "UNPHURL_API_KEY": "uph_your_key_here"
      }
    }
  }
}

2. No account yet?

The AI can create one for you. Just ask it to sign up for Unphurl. The signup tool works without an API key. After signup, add the key to your MCP configuration and restart.

3. Optional: custom API URL

For local development or staging, set UNPHURL_API_URL in the env block. Defaults to https://api.unphurl.com.

Tools

Tool	Description	Auth
signup	Create a new account, get an API key	No
resend_verification	Resend verification email (rate limited 3/hour)	Yes
check_url	Check a single URL for security/quality signals	Yes
check_urls	Batch check up to 500 URLs (handles async polling automatically)	Yes
list_profiles	List custom scoring profiles	Yes
create_profile	Create or update a scoring profile	Yes
delete_profile	Delete a scoring profile	Yes
show_defaults	Show all 23 scoring signals with default weights	No
get_balance	Check pipeline check credit balance	Yes
get_stats	View usage statistics and score threshold counts	Yes
get_pricing	Show available credit packages and pricing	No
purchase	Purchase credits (returns Stripe Checkout URL)	Yes
check_history	View recent URL check history	Yes

How billing works

• A positive credit balance is required for all check requests, even for free lookups (Tranco and cached domains). Free lookups don't deduct credits, but you need at least 1 credit on your account.
• Batch checks (check_urls): credits for all unknown URLs are deducted upfront. If you don't have enough credits for every unknown in the batch, the entire batch is rejected. You still get a smart summary showing how many URLs are known, cached, and unknown, and exactly how many credits you need to purchase.
• Failed pipeline checks are automatically refunded. If the pipeline fails completely and returns zero signals, the credit is restored. Partial results (some signals succeeded, some timed out) are not refunded since you received intelligence.

Filtering batch results

After a batch check with check_urls, ask your AI to filter the results:

"Check these 500 URLs. Give me two lists: the clean ones (score under 25) and the flagged ones (score 50 or higher). Export both as CSV."

The agent gets the batch results, filters by score, and outputs the lists you need. No code, no jq, no scripting. This is one of the advantages of the MCP path: your AI is the filter.

Companion skill

The check-url-safety skill teaches your AI to proactively check URLs before following or recommending them, without being asked.

If you use Claude Cowork or Claude Desktop: Just ask Claude: "Install the Unphurl URL safety skill." Claude can set it up for you.

If you use Claude Code, Cursor, or another developer tool:

# Global (all projects)
cp node_modules/@unphurl/mcp-server/skills/check-url-safety.md ~/.claude/skills/

# Or for a specific project
cp node_modules/@unphurl/mcp-server/skills/check-url-safety.md .claude/skills/

The skill teaches the agent to:

• Check unfamiliar URLs before recommending them
• Use batch checking for multiple URLs (more efficient)
• Handle zero-credit situations gracefully
• Skip well-known domains that don't need checking

Links

• API Documentation
• CLI Tool
• Website

License

MIT