Diffbot MCP Server
Extract, search, and map structured web data through Diffbot Knowledge Graph tools.
Data last scanned today · Reviewed today
Overview
This MCP server provides a suite of tools for turning unstructured web content into reliable data. By integrating with the Diffbot Extract API and Knowledge Graph, it enables AI agents to perform complex tasks like entity resolution, site crawling, and targeted web searches without the bloat of raw HTML.
Our verdict
This server carries a high risk level with a security score of 35/100. It is a community-maintained project rather than an official vendor release and lacks an editorially reviewed authentication process. It should be treated as experimental software and is not recommended for environments requiring strict security vetting.
- Supports extraction into meaningful and consistent JSON structures
- Contains zero direct dependencies to minimize supply-chain risk
- Provides NLP tools for identifying and resolving proper nouns
- Features built-in token management to optimize LLM usage
- Security score of 35/100 indicates significant high-risk factors
- Authentication method remains unreviewed by security standards
- Lacks official maintenance by the Diffbot vendor
Tools
| Tool | Description | Risk |
|---|---|---|
| extract | Web fetch tool that classifies the page and extracts content into a meaningful and consistent JSON structure. | low |
| search_web | Web search tool that ranks accuracy and returns markdown formatted content chunks. | low |
| enhance | Web lookup tool for returning structured data from the Diffbot Knowledge Graph on an organization or person. | low |
| resolve_entities | Recognize and resolve organizations, people, places, addresses, and other proper nouns in text to entity IDs. | low |
| crawl | Find all links on a web page and extract them all. | low |
| dql | Queries Diffbot Knowledge Graph for organization, person, article, place, and product entities. | low |
| dql_ontology | Helper utility for accurate dql ontology navigation. | low |
Compatibility
| Client | Local | Docker | Remote | Read-only |
|---|---|---|---|---|
| Claude Desktop | ||||
| Cursor | ||||
| VS Code | ||||
| Windsurf | ||||
| ChatGPT |
Frequently asked questions
›What is required to authenticate with this MCP server?
You must provide a valid Diffbot token, which can be obtained for free through the Diffbot website.
›Is it possible to use this server without installing it locally?
Yes, you can connect your MCP client directly to the remote server hosted by Diffbot using the URL https://mcp.diffbot.com/mcp?token=<YOUR_DIFFBOT_TOKEN>.
›How does the extract tool differ from a standard web fetcher?
The extract tool classifies web content and converts it into a consistent JSON structure, which helps minimize token usage compared to fetching raw HTML or markdown.
›What is the purpose of the dql_ontology tool?
This utility provides reference information for Diffbot's ontology and taxonomy, enabling the server to construct accurate DQL queries.
›Can I run this server using Docker?
Yes, you can build a Docker image using the provided Dockerfile and configure your MCP client to run it as a container.
Alternatives
Badge
Maintain this server? Add the live badge to your README.