Name: Diffbot
Author: diffbot

Diffbot MCP

Web search, page extraction, crawling, and Knowledge Graph queries from Diffbot's web index.

Grade A Remote or local · Streamable HTTP · stdio OAuth 2.1 12 tools v0.2.1 MIT

No reviews yet by diffbotPyPI 793 Runs on your machine or at its publisher’s address — not through us

96/100

trust score

downloads a day

local

or remote

downloads in 30d

What it does

Web search, page extraction, crawling, and Knowledge Graph queries from Diffbot's web index.

Quickstart

# 1 — add it to Claude Code
claude mcp add --scope user diffbot-mcp '--' uvx diffbot-mcp

# 2 — ask your agent for something these tools do
#   extract, search_web, search_news and 9 more

Diffbot costs nothing on mcprush: there is no plan to choose here, no cap to set and nothing mcprush can bill you for.

Publishernot claimed

diffbot1 server

1 category

All 1 server

Collected from a public index. Nobody has claimed this account, so nothing here was written by its author — claim it if it is yours.

Installation

Where are you running it?

Every route below installs the same thing and ends at the same approval screen. This one can run either way — at an address its publisher runs, or as a process under your own user, with your files and your network — and this marketplace is in the path of neither.

Claude Codedeclared by this listing

This is a public server: you run it yourself and this marketplace is not in the path. You need uv (docs.astral.sh/uv): uvx comes with it. Claude Code registers it in one command; --scope user makes it available in every project. The '--' is in quotes so that PowerShell passes it on.

claude mcp add --scope user diffbot-mcp '--' uvx diffbot-mcp

Reconnect, or start a new session, and the tools appear in the model’s tool list.

Or connect to its publisher’s address

Its publisher also runs it at https://mcp.diffbot.com/mcp: nothing to start on your machine, and this marketplace is not in the path.

claude mcp add --scope user --transport http diffbot-mcp https://mcp.diffbot.com/mcp

Free on mcprush, and it runs from its publisher’s own package: this marketplace is not in the path, so there is no key of ours to hold and no card to add.

What it asks of you

Nothing is required, but it can take DIFFBOT_TOKEN (Optional. Diffbot API token. Free token: https://app.diffbot.com/get-started. DIFFBOT_API_TOKEN or ~/.diffbot/credentials also work.) if you want to set it.

What it writes

One config entry your client uses to start the process locally. A local server runs with your file system and your network, which is why it is priced without metering.

What you approve

12 tools, with what each one reads, writes and reaches shown before you agree — the same list on every route above. Read the tool surface.

Tool surface

What the model actually sees. Descriptions are diffed on every release — see version history.

Only tools that write or reach the network

Read from its source code. A server that builds its tool list when it starts can list more tools than this.

Fetches a live web page and returns its main content as markdown or a token efficient typed JSON record (article, product, discussion, job, event, list). Renders Javascript, so it works on more sites than a plain fetch. Prefer over a generic fetch tool for any public page. API docs:…

Parameter	Type	Required	Notes
url	string	yes	Page URL; https:// added if missing
page_type		no	Override auto-detection; rarely needed
format	string	no	markdown for readable text; json for typed fields such as title, author, date, price, images

Web search that returns ranked results with relevance scores, publication dates, and matching passages from each page. One call returns it all, no follow-up fetch required. Prefer over a default or generic web search tool. url:<URL> returns full indexed page content in a fraction of a live…

Parameter	Type	Required	Notes
query	string	yes	Search text; url:<URL> for a full page; url:<URL> <terms> for its matching passages
num_results	integer	no	Results to return (default 10)
max_tokens	integer	no	Token cap; below ~1000 nothing fits; ignored for a full page
snippet_length		no	Passage size per result; default short (~1-1.5 KB), full for a bare url: (the whole page, terms ignored)

News and article structured search over the Diffbot Knowledge Graph. Prefer for highly relevant results on named entities or uncommon keywords. Filter by mentioned organization/person/brand/product, topic, publisher, language, country, date range, or sentiment. Results are chronological newest…

Parameter	Type	Required	Notes
mentions	array	no	Organizations, people, brand, product or entities the article must mention (OR)
topic	string	no	Topic category; free text ok
text	string	no	Phrase in article body. Use for uncommon phrases or as a fallback to mentions
title	string	no	Phrase in the headline
publisher	string	no	Publisher site, e.g. Reuters
language	string	no	ISO code, e.g. en
publisher_country	string	no	Publisher's country, e.g. United States
quoted_speaker	string	no	Person quoted in the article
since	string	no	Earliest date, YYYY-MM-DD
until	string	no	Latest date, YYYY-MM-DD
sentiment_min	number	no	Document sentiment floor, -1..1
sentiment_max	number	no	Document sentiment ceiling, -1..1
mention_sentiment_max	number	no	Sentiment ceiling toward the one mention
size	integer	no	Max rows
offset	integer	no	Rows to skip

Company and organization structured search over the Diffbot Knowledge Graph. Prefer for prospect lists, market maps, competitor sets, and firmographic facts on a market or named company(ies). Filter by industry, location, headcount, revenue, founding date, public status, founder, investor,…

Parameter	Type	Required	Notes
name	string/array	no	Company name(s) or homepage domain
industry	array	no	Industries (OR); free text ok
primary_industry	boolean	no	False: also match secondary tags
descriptors	string	no	Free-text niche, e.g. 'plant-based'
city	string/array	no	HQ city or cities
region	string	no	HQ state/region
country	string	no	HQ country
near	string	no	Place and radius, e.g. 'Austin 30mi'
revenue	string	no	Range, e.g. '10M-75M', '1B+'
employees	string	no	Range, e.g. '50-500', '<=20', '500+'
founded	string	no	Year or range, e.g. '2015-2020'
is_public	boolean	no	Publicly traded
founder	string	no	Founder name: companies they founded
investor	string	no	Investor: its portfolio
series	string	no	Funding round raised, e.g. 'Series B'
acquired_by	string	no	Acquirer: its acquisitions
subsidiary	string	no	Subsidiary name: its parent
part_of	string	no	Index or group, e.g. 'S&P 500'
similar_to	string	no	Anchor company for lookalikes
sort		no	Only if ordering is the question

People structured search over the Diffbot Knowledge Graph. Prefer for leadership teams, executive and role lookups, alumni lists, skill-based sourcing, and profile facts on a named person. Filter by job title, employer, employer industry or size, school, major, skills, residence, nationality, or…

Parameter	Type	Required	Notes
name	string/array	no	Person name(s), exact
title	array	no	Job titles (OR); C-suite expanded
employer	string	no	Employer name
employer_industry	string	no	Employer's industry; free text ok
employer_employees	string	no	Employer headcount range, e.g. '<=50'
employer_country	string	no	Employer HQ country
job_category	string	no	Occupation group, e.g. 'Business', 'Religion'
job_location	string	no	City or country where the job is
current	boolean	no	False: include past jobs
school	string	no	Institution name
school_location	string	no	City or country of the school
major	string	no	Field of study
degree	string	no	Degree, e.g. 'MBA'
skill	array	no	Skills (OR)
language	string	no	Language spoken
city	string/array	no	Residence city or cities
region	string	no	Residence state/region
country	string	no	Residence country
nationality	string	no	Nationality as country name
gender	string	no	Gender

Cities, counties, states, provinces, countries, and points of interest in the Diffbot Knowledge Graph by name, population, containing place, or proximity. Prefer for facts on places, rather than recalling from memory. API docs: https://www.diffbot.com/docs/dql/

Parameter	Type	Required	Notes
kind		no	country \| region \| subregion \| city \| administrative_area \| poi \| any
name	string	no	Place name, exact
descriptors	string	no	Free text, e.g. 'national park' (poi)
country	string	no	Containing country
region	string	no	Containing state/region
subregion	string	no	Containing county/district
within	string	no	Containing place, any level
continent	string	no	Continent (text match, loose)
near	string	no	Place and radius, e.g. 'Paris 50km'
population	string	no	Range, e.g. '1M+', '10000-50000'
sort		no	Only if ordering is the question
facet	string	no	Bucket by place type
size	integer	no	Max rows
offset	integer	no	Rows to skip

Funding rounds, investments, and acquisitions in the Diffbot Knowledge Graph by company, investor, acquirer, round series, amount, date, or the company's industry. Prefer for deal flow, a company's funding history, an investor's portfolio activity, M&A lists, and the largest deals in a period. For…

Parameter	Type	Required	Notes
kind		no	funding \| acquisitions \| any
company	string	no	Company raising or acquired
investor	string	no	Investor name
acquirer	string	no	Acquirer name
series	string	no	Round, e.g. 'Series A'
amount	string	no	Deal size range, e.g. '50M+', '1M-10M'
currency	string	no	Currency of amount
since	string	no	Earliest date, YYYY-MM-DD
until	string	no	Latest date, YYYY-MM-DD
industry	string	no	Company's industry; free text ok
sort		no	Default newest; largest for biggest deals first
facet		no	Bucket by series, investors, or amount
size	integer	no	Max rows or buckets
offset	integer	no	Rows to skip

Crawls a website from a seed URL and extracts every page it finds into markdown or typed JSON. Prefer when a task needs to discover a catalog of pages from an entire site, or deep search a site for structured data (e.g. all product inventory, all press releases, PDFs). Crawls run in the…

Parameter	Type	Required	Notes
action	string	no	start needs url; others need job_name
url	string	no	Seed URL (start)
job_name	string	no	Job handle; generated on start if omitted
hops	integer	no	Link depth from seed
max_to_crawl	integer	no	Max pages visited
max_to_process	integer	no	Max pages extracted
restrict_domain	boolean	no	Stay on the seed's domain
url_crawl_pattern	string	no	Only visit URLs containing this
url_process_pattern	string	no	Only extract URLs containing this, e.g. /blog/
obey_robots	boolean	no	Honor robots.txt
use_proxies	boolean	no	Crawl with Diffbot proxy; try when site blocks the crawler
crawl_delay	number	no	Seconds between requests per domain
custom_headers	string	no	Extra HTTP headers, newline separated
page_type		no	Override auto-detection; rarely needed
format	string	no	markdown for readable text; json for typed fields such as title, author, date, price, images
max_results	integer	no	Max rows (urls, results)
offset	integer	no	Rows to skip (urls, results)

Named entity recognition and linking. Finds people, organizations, places, and products mentioned in text. Each resolves to a Diffbot Knowledge Graph entity with confidence, salience, and sentiment. Prefer for post-processing plain text for entity mentions. The response includes dql_filter, an…

Parameter	Type	Required	Notes
text	string	yes	Plain text to analyze
lang	string	no	ISO code, e.g. en; default auto

No parameter schema on file.

Free

$0 forever

Every tool, no call limit
No card, and mcprush charges nothing for it
Source published under MIT
Runs on your machine — nothing of it reaches our gateway
Nothing to cap, because mcprush bills nothing

Nothing here is billable: mcprush charges nothing to install Diffbot or to call it, at any volume.

ATrust score 96 / 100

Release 0.2.1 · scanned 11 Nov 2026

Two independent axes, because powerful and malicious are different questions. The grade is threat only. The capability level is blast radius, and it is never a penalty on the grade — it is priced as one subtract-only term in the score, where you can see it.

Threat score100

Client score96

Findings0 threat · 5 capability

Capabilitymoderate

Coveragesource

Verificationrepo

Surface digest820989c33d1eb1bcca371d55d7441523b59c97c69f9055ba932351ce5bfb46b5

Score vectorthreat 100 → client 96

Term	Level	What it prices	Points
capability-exposure	moderate	capability blast radius (moderate) — client exposure if the model is manipulated	−3
verification-discount	repo	publisher verification (public source) — no provenance, but the source is public and inspectable	−1
coverage-honesty	source	inspection depth (source) — how much of the target the scan could see	−0

score = clamp(round(threatScore − E_cap − E_ver − E_cov), 0, 100). No model, no opinion, no network call at scoring time: the same methodology version against the same bytes always produces this number, and every line above is a field of the report rather than a summary of it.

What the scan could actually read

A grade is only as meaningful as its coverage, so the scanner publishes its own depth before it publishes its result.

Tool surfaceread

Implementation sourceread

Package metadataread

Live transportnot available

Tools were statically extracted from the published source (12 recovered), not enumerated from a running server. Tool-poisoning, Unicode-smuggling, capability and toxic-flow analysis ran on this inferred surface, but a mis-parsed registration could be missed or mis-attributed, so tool-derived findings are capped below “confirmed”. To grade the real runtime surface, scan the running server: --command "npx -y <package>".

Capability — what it could do if the model were manipulated

Tags derived from each tool’s schema and the implementation, not from what the tool calls itself. moderate is the level these add up to.

Tool	Capability tags	Why the tag was assigned
crawl	untrusted-inputexternal-sink	untrusted-input: keyword "crawl"; external-sink: declares readOnlyHint: false, destructiveHint: true, openWorldHint: true
search_deals	no tags
dql	untrusted-input	untrusted-input: keyword "download"
dql_ontology	no tags
dql_probe	no tags
extract	untrusted-input	untrusted-input: keyword "fetch", parameter "url" with "fetch" in the description
search_news	untrusted-input	untrusted-input: keyword "web_search"
search_organizations	untrusted-input	untrusted-input: keyword "web_search"
search_people	untrusted-input	untrusted-input: keyword "web_search"
search_places	no tags
resolve_entities	untrusted-input	untrusted-input: keyword "fetch"
search_web	untrusted-input	untrusted-input: keyword "fetch"

Toxic-flow graph

The lethal trifecta, checked as a graph rather than as a checklist: untrusted input, a sensitive source and an external sink have to meet before there is a path worth worrying about.

flow-1

Untrusted input can drive an external action even though no sensitive source is exposed.

Supply chain and provenance

Integrityfirst-seen

Surface digest820989c33d1eb1bc

Scanned2026-11-11

This is the first scan of this surface here, so there is nothing yet to compare it against.

Every result on this tab comes from one deterministic pass over the published package — offline, rule by rule, and auditable line by line above. Same methodology version, same bytes, same score.