urlDNA MCP Server

Overview

The urlDNA MCP server provides an AI‑ready interface for deep URL analysis, brand monitoring, and phishing detection. It exposes a set of RESTful endpoints that let Claude (or any MCP‑compatible assistant) submit URLs for scanning, retrieve detailed results, perform targeted searches, and quickly check whether a link is safe or malicious. By packaging these capabilities into an MCP server, developers can embed advanced web‑security intelligence directly into conversational agents without managing the underlying API calls themselves.

Problem Solved

Modern applications increasingly rely on external links—whether in marketing campaigns, user‑generated content, or internal references. Ensuring every URL is trustworthy and complies with brand guidelines can be laborious, especially at scale. The urlDNA server automates this process: it gathers comprehensive metadata (HTML content, HTTP headers, visual screenshots across multiple viewports), evaluates malicious indicators, and indexes data for rapid search. This eliminates manual vetting, reduces security risk, and keeps brand consistency across platforms.

Core Capabilities

• Create Scan – Initiate a full scan of any URL, optionally specifying device type, viewport size, and user‑agent string. The server returns a scan ID that can be polled for results.
• Get Scan – Retrieve the complete analysis report, including content classification, malware flags, and visual snapshots for each requested viewport.
• Search – Query the indexed dataset using a custom query language (e.g., ) to surface relevant scans or historical data.
• Viewports & User Agents – List all supported device configurations and browser fingerprints, enabling precise simulation of real‑world browsing conditions.
• Fast Check (Beta) – A lightweight endpoint that instantly classifies a URL as CLEAN or MALICIOUS, ideal for real‑time filtering in chat flows.

These features are exposed through straightforward MCP methods such as , , and , allowing an AI assistant to request a scan, wait for completion, and present the findings in natural language.

Use Cases

• Content Moderation – An assistant can automatically flag or block links that fail the fast check before a user posts them in a forum or messaging app.
• Brand Monitoring – By searching for domain patterns, the server can detect unauthorized use of a brand’s name or logos across the web.
• Phishing Detection – The scan results include malware signatures and suspicious patterns, enabling proactive alerts to security teams.
• SEO & Analytics – Detailed metadata from scans can feed into SEO dashboards or compliance reports, giving stakeholders insight into link performance and safety.

Integration with AI Workflows

Because the server follows the MCP specification, Claude can invoke its methods as tool calls within a conversation. For example, after a user asks whether a link is safe, the assistant can call , receive an immediate verdict, and explain the reasoning. For deeper investigations, it can request a full scan, wait for completion, and then summarize the findings. The server’s search capability also lets the assistant pull historical data to support trend analyses or evidence gathering.

Unique Advantages

• Comprehensive Visual Analysis – By capturing screenshots across multiple viewports, the server offers richer context than simple textual scans, helping users spot visual phishing tricks.
• Custom Query Language – The flexible search syntax allows developers to craft precise filters without exposing raw API endpoints.
• Beta Fast Check – Provides near‑instant safety checks, reducing latency in conversational interactions where speed is critical.
• Python SDK Ready – While the MCP server abstracts away implementation details, developers can still leverage the existing Python package for non‑MCP use cases or testing.

In summary, the urlDNA MCP server equips AI assistants with powerful web‑security intelligence, streamlining URL vetting, brand monitoring, and threat detection directly within conversational workflows.