MCP Server

The Scrapio MCP server gives AI agents direct access to the full web data surface through a single registered server — fetch, crawl, Google and Fast Search, YouTube, Amazon and Walmart product data, Booking.com and Agoda hotel data, Airbnb and Target listings, Reddit, TikTok, and Apple App Store data, Bing search, ChatGPT/Perplexity/Gemini prompting, browser automation, and async jobs. Also listed on the official MCP registry as io.github.xsronhou/scrapio-mcp.

Install

No installation required. Run it with npx:

SCRAPIO_API_KEY=sk-... npx @scrapio/mcp

The server starts on stdio by default and is ready to accept connections from your agent runtime.

Configure your agent

Hosted (no install)
HTTP (remote / self-hosted)

Point your agent runtime directly at Scrapio’s hosted MCP endpoint — nothing to install or run locally:URL: https://mcp.scrapio.dev/mcp

OAuth (recommended)

For clients with a “paste a URL” connector flow (Claude Desktop’s Settings → Connectors → Add custom connector, and similar), just enter the URL above. You’ll be redirected to log in to your Scrapio account and approve access — no API key to find or paste. Each login/approval is scoped to your own account; the hosted endpoint is multi-tenant, so different accounts never see each other’s sessions or data.Behind the scenes: the client registers itself automatically (no setup on your end), then runs a normal OAuth authorization-code flow against https://mcp.scrapio.dev. If your client doesn’t support this, use the manual method below instead.

Manual API key

For clients that only support a static header (or if you’d rather not go through the login flow), send Authorization: Bearer <SCRAPIO_API_KEY> on every request — for the JSON-based configs above, add a "headers" block instead of "env":

{
  "mcpServers": {
    "scrapio": {
      "url": "https://mcp.scrapio.dev/mcp",
      "headers": {
        "Authorization": "Bearer sk-..."
      }
    }
  }
}

This is the same underlying server and the same per-tenant isolation as the OAuth method — just authenticated with a fixed key instead of a login flow.

For deployments where the MCP server runs as a persistent process rather than a child process:

SCRAPIO_API_KEY=sk-... npx @scrapio/mcp --transport http --port 3010

The server listens on http://localhost:3010/mcp. Configure your agent runtime to use this URL as the MCP endpoint. You are responsible for securing the HTTP endpoint at the network level (private VPC, reverse proxy with auth, etc.).

Available tools

Tool	What it does
`fetch`	Fetch a URL and return its content as markdown, HTML, JSON, or a screenshot
`google_search`	Search Google and return structured results — organic, news, images, or shopping
`fast_search`	Lightweight organic + news search, same price as `google_search` (15 credits), but no HTML capture and requires only a Starter plan instead of Pro
`interact`	Drive a real browser through a multi-step workflow — click, type, scroll, wait
`crawl`	Crawl one or more seed URLs and return content for every page visited
`submit_job`	Submit a long-running job asynchronously and return a job ID
`get_job`	Check job status and retrieve the result when complete
`amazon_product`	Get structured product data from Amazon by URL, ASIN, or search query. Requires a Starter plan or higher.
`amazon_search`	Search Amazon and return structured product listings. Requires a Starter plan or higher.
`amazon_search_crawl`	Queue an async, multi-page Amazon search crawl — returns a job ID. Requires a Starter plan or higher.
`walmart_search`	Search Walmart and return structured product listings. Requires a Starter plan or higher.
`walmart_search_crawl`	Queue an async, multi-page Walmart search crawl — returns a job ID. Requires a Starter plan or higher.
`youtube_transcript`	Get the full transcript for a YouTube video by URL or video ID. Requires a Starter plan or higher.
`youtube_search`	Search YouTube and return structured video results. Requires a Starter plan or higher.
`youtube_video`	Get structured metadata for a single YouTube video. Requires a Starter plan or higher.
`youtube_search_crawl`	Queue an async, multi-page YouTube search crawl — returns a job ID. Requires a Starter plan or higher. Capped at 700 total results (10 credits per page crawled).
`booking_search`	Search Booking.com for hotels by location and stay dates. Requires a Starter plan or higher.
`booking_property`	Get structured details — amenities, room rates, review score — for a single Booking.com property. Requires a Starter plan or higher.
`booking_reviews`	Get guest reviews and the score breakdown for a Booking.com property. Requires a Starter plan or higher.
`agoda_search`	Search Agoda for hotels by location and stay dates. Requires a Starter plan or higher.
`agoda_property`	Get structured details — amenities, room rates, review score — for a single Agoda property. Requires a Starter plan or higher.
`agoda_reviews`	Get guest reviews and the score breakdown for an Agoda property. Requires a Starter plan or higher.
`airbnb_search`	Search Airbnb listings for a free-text location. Requires a Starter plan or higher.
`airbnb_listing`	Get structured detail for a single Airbnb listing by id. Requires a Starter plan or higher.
`autotrader_search`	Search AutoTrader UK car listings near a postcode, with optional make/model/year filters. Requires a Starter plan or higher.
`autotrader_listing`	Get structured detail for a single AutoTrader UK listing by its numeric listing id. Requires a Starter plan or higher.
`target_search`	Search Target.com and return structured product results. Requires a Starter plan or higher.
`target_product`	Get structured product details from Target.com. Requires a Starter plan or higher.
`reddit_post`	Get a Reddit post and its top-level comments. Requires a Starter plan or higher.
`reddit_subreddit`	Get a subreddit’s post feed. Requires a Starter plan or higher.
`reddit_user`	Get a Reddit user’s mixed post/comment activity feed. Requires a Starter plan or higher.
`hacker_news_item`	Get a Hacker News item (story, comment, job, poll, or pollopt) by id, optionally with its comment tree. Requires a Starter plan or higher.
`hacker_news_user`	Get a Hacker News user’s profile — karma, about text, and submitted item ids. Requires a Starter plan or higher.
`hacker_news_search`	Search Hacker News stories and comments by keyword. Requires a Starter plan or higher.
`hacker_news_stories`	Resolve a Hacker News story-id list (top, new, best, ask, show, job) to full item details. Requires a Starter plan or higher.
`tiktok_post`	Get structured metadata for a TikTok video post. Requires a Starter plan or higher.
`apple_app_store_app`	Get structured metadata for an Apple App Store app listing. Requires a Starter plan or higher.
`bing_search`	Search Bing and return structured organic and paid results. Requires a Starter plan or higher.
`chatgpt_prompt`	Send a prompt to ChatGPT and return the structured response, including citations. Requires a Starter plan or higher.
`perplexity_prompt`	Send a prompt to Perplexity and return the structured, web-search-grounded response. Requires a Starter plan or higher.
`gemini_prompt`	Send a prompt to Gemini and return the structured response, including citations. Requires a Starter plan or higher.
`youtube_channel`	Get channel metadata and recent videos for a YouTube channel handle. Requires a Starter plan or higher.
`product_hunt_leaderboard`	Get the Product Hunt leaderboard: ranked products with tagline, upvotes, comment count, topics, and makers. Fast, but capped at the page’s initial render (`truncated: true`). Requires a Starter plan or higher.
`product_hunt_leaderboard_crawl`	Queue an async job that scrolls a Product Hunt leaderboard to its genuine end and returns a job ID. Requires a Starter plan or higher.
`product_hunt_product`	Get structured details for a single Product Hunt launch: name, tagline, description, topics, upvotes, comment count, makers, and website link. Requires a Starter plan or higher.
`product_hunt_maker`	Get a Product Hunt maker/founder profile: name, headline, follower count, and the products they’ve shipped. Requires a Starter plan or higher.

Example prompts

Once the server is connected, your agent can use natural language to call these tools:

“Fetch the content of https://news.ycombinator.com and summarize the top 5 stories.”

“Search Google for ‘best TypeScript ORM 2025’ and give me the top 3 results.”

“Get the transcript of this YouTube video: https://www.youtube.com/watch?v=dQw4w9WgXcQ”

“Find the current price of the Sony WH-1000XM5 headphones on Amazon.”

“Crawl https://docs.example.com up to 20 pages and extract all the API endpoint descriptions.”

“Go to https://app.example.com, click the login button, type my email and password, and return the dashboard content.”

“Search Booking.com for hotels in Lisbon, check-in Sept 10, check-out Sept 14, for 2 adults, sorted by rating.”

“Search Agoda for hotels in Bangkok under $100/night and show me the top 5 with their review scores.”

“Get the room rates and amenities for this Booking.com property: https://www.booking.com/hotel/pt/example.html”

Async jobs

Some operations — large crawls, slow pages, multi-step interactions — take longer than a single tool call can wait. Use submit_job to queue the work and get_job to poll for the result:

Agent: submit_job(kind="crawl", input={ seeds: ["https://docs.example.com"], max_pages: 50 })
→ { job_id: "abc123", status: "queued" }

Agent: get_job(job_id="abc123")
→ { status: "running", ... }

Agent: get_job(job_id="abc123")
→ { status: "completed", result: { pages: [...] } }

Environment variables

Variable	Required	Description
`SCRAPIO_API_KEY`	Yes, unless using the hosted endpoint	Your API key. Get one at scrapio.dev. Not used by the hosted endpoint (`mcp.scrapio.dev`), which takes the key per-request via the `Authorization` header instead.
`SCRAPIO_BASE_URL`	No	Override the API base URL (for local development or staging).
`MCP_PORT`	No	Port for HTTP transport mode. Defaults to `3010`.

Troubleshooting

The server starts but no tools appear in my agent

Restart your agent runtime after adding the server config — most runtimes only discover MCP servers on startup.

Error: SCRAPIO_API_KEY is not set

The env block in your agent config is not being passed to the process. Double-check that your config file uses the correct format for your runtime and that the key name matches exactly.

Tool calls return 'Authentication failed'

Your API key is set but invalid or revoked. Generate a new key in the dashboard.

Tool calls return 'Credits exhausted'

Your account has no remaining credits. Add credits at app.scrapio.dev.

npx is slow to start

The first run downloads the package. Subsequent calls use the npm cache and start in under a second. For production deployments, install globally with npm install -g @scrapio/mcp and reference the binary directly.

A tool call suddenly fails with a generic error, after previously working

If you’re connected to the hosted endpoint (mcp.scrapio.dev), this usually means the server was redeployed and your client’s existing connection is stale. Disconnect and reconnect the connector (for Claude Desktop: Settings → Connectors → remove and re-add, or toggle it off and on) to establish a fresh session. This doesn’t affect the local npx/stdio setup, which reconnects automatically on your runtime’s next restart.

Getting Started

Guides

API Reference

Install

Configure your agent

OAuth (recommended)

Manual API key

Available tools

Example prompts

Async jobs

Environment variables

Troubleshooting

​Install

​Configure your agent

​OAuth (recommended)

​Manual API key

​Available tools

​Example prompts

​Async jobs

​Environment variables

​Troubleshooting

Install

Configure your agent

OAuth (recommended)

Manual API key

Available tools

Example prompts

Async jobs

Environment variables

Troubleshooting