Model Context Protocol

Give your AI agent
the data it needs

lobstr.io's scrapers — Google Maps, LinkedIn, e‑commerce and hundreds more — now plug straight into Claude, ChatGPT and any MCP client. Your agent picks a scraper, runs it, and brings back the rows. You just ask.

https://mcp.lobstr.io/mcp
Sign in with lobstr.io — no API key 22 tools · read & run · credit‑safe
Works in ClaudeChatGPTCursorGrokVS Code

Try:

Thousands of teams use lobstr.io to collect web data at scale — now your AI agent can too.

Connect in a minute

Three steps to a data‑savvy agent

No SDK, no glue code. Add one URL, sign in with lobstr.io, and start asking.

1

Add the server

Drop the lobstr.io URL into your client's connectors — in Claude, Settings → Connectors → Add custom connector.

https://mcp.lobstr.io/mcp
2

Sign in with lobstr.io

A quick browser approval on lobstr.io. You choose what the agent may do — reading is free, running spends credits.

✓ approve · no token to copy
3

Just ask

Tell your agent what data you want. It finds the right scraper, runs it, and hands back the results.

"Get dentists in Manchester"

Using a config‑file client? Show the JSON snippet ↓  ·  Full setup for Claude Code, Cursor, ChatGPT, Codex & Grok is in the docs →

The toolbox

Everything your agent can do

Twenty-two tools, grouped by what they touch. Reading is always safe — running a scraper estimates the cost and asks you first, so nothing spends credits by surprise.

Discover scrapers

Read

Search lobstr.io's catalog, read a scraper's inputs, output fields and pricing, and list the ones you've already saved.

search_scrapers get_scraper_details list_my_scrapers get_my_scraper

Track runs

Read

Check the status and progress of any run, wait for one to finish — and list the recent runs for a scraper.

get_run wait_for_run list_runs

Read results

Read

Pull a page of scraped results keeping just the fields you need — or grab a link to the full dataset.

get_results get_results_url

Your account

Read

Check your lobstr.io credit balance and see who you're signed in as.

check_credits whoami

Connected accounts

Free

See the platform accounts (LinkedIn, Facebook, …) your scrapers log in with and their status, and link one to a scraper — metadata only, never credentials.

list_accounts get_account attach_account

Build a run

Free

Create a scraper without running it, add input rows in batches, and see the credit cost and time estimate before anything starts.

create_squid add_tasks estimate_run

Run & manage

Spends credits

Launch a scraper on your behalf — it estimates the cost and waits for your OK first. Change a saved scraper's settings, stop a run, clear its tasks to reuse it, or free its slot (those are free).

run_scraper update_scraper abort_run empty_scraper deactivate_scraper

Get the data you need — now inside your AI agent

Point Claude or ChatGPT at lobstr.io and let it do the collecting. Free to connect; you only spend credits when you run a scraper.

Open dashboard

Server URL  ·  https://mcp.lobstr.io/mcp