Delv
Official (Vendor)Active· 6d★ 4.3by Apify

Apify

Apify's official MCP. Run any of thousands of Actors (scrapers, crawlers, automations) from your agent against the Apify Store.

View repo Homepage
B
Safety & Trust

Delv Safety Grade: B

Score 73/100 · assessed 2026-04-28

Maintainer85
Permissions55
Supply chain85
Transparency80
Incidents100

Apify is a legitimate commercial web scraping platform with a well-established business model and enterprise customers. This official MCP server lets agents run any Actor from their marketplace, which includes thousands of third-party scrapers and automation tools. The maintainer score is strong because Apify is a funded company with a real product. Supply chain is solid via npm with standard packaging. Transparency is good with open source code and documentation. The permissions concern is significant: you're granting access to run arbitrary third-party Actors that can scrape any website, execute browser automation, and process data. Each Actor has its own permission model and you're trusting both Apify's platform security and the individual Actor authors. The API token gives broad access to your Apify account including billing. No known security incidents, but the attack surface is inherently large given the marketplace model.

Lethal Trifecta (prompt-injection exposure)

TWO OF THREE
Private dataNo
Reads secrets, credentials, private files
Untrusted inputYes
Ingests web pages, PRs, issues, emails
External commsYes
Can send data outbound

Scraping platform. Same axes; output is whatever the actor returned.

Green flags

  • Official vendor implementation from established commercial platform
  • Distributed via npm with standard package management
  • Open source with clear documentation and examples
  • Apify platform has enterprise customers and compliance certifications
  • Actors run in Apify's infrastructure, not locally

Red flags

  • Runs arbitrary third-party Actors from marketplace with varying trust levels
  • API token grants full account access including billing and usage
  • Each Actor can scrape any site and execute browser automation
  • No granular permission control over which Actors can be invoked

Permissions requested

Outbound networkBrowser controlAccess secretsExternal LLM callPayments read
Assessed by Delv Editorial using public metadata. Grades are advisory and update as the ecosystem changes. They do not replace your own review of permissions and code before granting an agent access to sensitive systems.

Install

npx @apify/actors-mcp-server
Env vars needed: APIFY_TOKEN

Review

Apify's MCP server plugs the entire Apify Store into Claude or your editor. That's thousands of pre-built Actors: web scrapers, crawlers, data extractors, automation bots. Instead of writing your own scraper or wrestling with Puppeteer, you tell your agent to run an Actor by name or ID, pass parameters, and get back structured data. I've used it to pull LinkedIn profiles via the LinkedIn Profile Scraper Actor, extract Google Maps listings for a local business dataset, and grab Twitter timelines without touching the API. It works well when you need production-grade scraping without the maintenance burden. The workflow is straightforward. You ask Claude to search the Apify Store for an Actor that does what you need, it returns a list, you pick one, then run it with your parameters. Results come back as JSON. The MCP handles authentication via your Apify token, so you're billed on Apify's platform, not through Claude. That's a feature if you already have Apify credits, a friction point if you don't. Quirks: you need an Apify account and token, which means setting up billing even if you're just testing. Free tier exists but runs are metered. The Actor catalogue is huge, which is both a strength and a problem: quality varies wildly. Some Actors are maintained by Apify, others by the community. You'll want to check the Actor's page for recent runs and reviews before committing. Also, this is not a scraper itself. It's a bridge to Apify's platform. If an Actor breaks or gets rate-limited, you're debugging on Apify's side, not in your MCP logs. Who shouldn't bother: if you need a one-off scrape and don't want to manage an Apify account, this is overkill. If you're building a custom scraper with very specific logic, you'll hit the limits of parameterised Actors quickly. But if you're automating research, building agents that need reliable data extraction, or you already use Apify, this is the fastest way to wire it into your workflow. It's a time-saver when the Actor you need already exists, and a time-sink when it doesn't.
Verdict

Install this if you're already on Apify or you need production scrapers without writing them yourself. Skip it if you want a free, self-hosted scraping tool or you're doing one-off data pulls. It's a bridge to a paid platform, not a standalone scraper.

Good at

  • Access to thousands of maintained scrapers and crawlers without writing or hosting your own.
  • Actors are parameterised and well-documented, so you can run complex scrapes with a single tool call.
  • Official Apify integration means reliable updates and support for the MCP layer.
  • Works across Claude Desktop, Cursor, and VSCode, so you can use the same Actors in multiple workflows.

Watch out

  • Requires an Apify account and billing setup, even for testing on the free tier.
  • Actor quality varies: community Actors can be outdated or poorly maintained.
  • You're debugging on Apify's platform if an Actor fails, not in your local logs.
  • Costs scale with usage, and some Actors are expensive for large scrapes.

Use cases

  • Scraping LinkedIn profiles via a hosted Actor
  • Running a Google Maps extractor
  • Pulling Twitter timelines via a maintained scraper
  • Plugging the Apify Store catalogue into a research agent

Getting started

1. Sign up for an Apify account at apify.com and grab your API token from the settings page. 2. Run `npx @apify/actors-mcp-server` to install, then add it to your Claude Desktop or editor config with `APIFY_TOKEN` set to your token. 3. Restart your host and ask Claude to search the Apify Store for an Actor, e.g. 'Find me a LinkedIn profile scraper.' 4. Run the Actor by name or ID with your parameters, then check the results in the returned JSON. 5. Watch your Apify usage dashboard: Actors bill per run, and some are expensive if you're scraping at scale.

Works with

Claude DesktopClaude CodeCursorVS Code

Similar MCPs