zrk-scrapers-mcp
# zrk-scrapers-mcp
An MCP server that gives AI agents (Claude, Cursor, VS Code, Windsurf and others) 9 production web scrapers. Each tool runs a maintained [Apify](https://apify.com/zrk1212) actor and returns clean JSON.
| Tool | What it gets | Actor |
|---|---|---|
| `scrape_ats_jobs` | Open jobs from Greenhouse, Lever, Ashby, SmartRecruiters, Workable, Recruitee, Personio, Teamtailor, Rippling | [ats-jobs-monitor](https://apify.com/zrk1212/ats-jobs-monitor) |
| `scrape_workday_jobs` | Every job on a Workday career site, past the 2,000 cap | [workday-jobs-scraper](https://apify.com/zrk1212/workday-jobs-scraper) |
| `audit_pagespeed` | Lighthouse scores, Core Web Vitals and top fixes for many pages or whole sites | [pagespeed-bulk-auditor](https://apify.com/zrk1212/pagespeed-bulk-auditor) |
| `find_uk_new_companies` | Newly incorporated UK companies with directors (Companies House) | [uk-new-company-leads](https://apify.com/zrk1212/uk-new-company-leads) |
| `get_tiktok_transcripts` | TikTok video transcripts from URLs or profiles | [tiktok-transcript-scraper](https://apify.com/zrk1212/tiktok-transcript-scraper) |
| `find_luma_organizers` | Luma event organizer leads with socials and audience size | [luma-organizer-leads](https://apify.com/zrk1212/luma-organizer-leads) |
| `get_google_hotel_prices` | Google Hotels prices from every seller for exact dates | [google-hotels-prices-scraper](https://apify.com/zrk1212/google-hotels-prices-scraper) |
| `scrape_google_ads_transparency` | A brand's Google ads with Search ad copy as text | [google-ads-transparency-ad-copy](https://apify.com/zrk1212/google-ads-transparency-ad-copy) |
| `scrape_telegram_channel` | Public Telegram channel posts, views, reactions and subscriber counts | [telegram-channel-scraper-stats](https://apify.com/zrk1212/telegram-channel-scraper-stats) |
| `get_run_results` | Results of a long run, or the next page of a big one | |
## Setup
1. Get an Apify API token (free account, includes monthly credit): https://console.apify.com/settings/integrations
2. Add the server to your client.
**Claude Code**
```bash
claude mcp add zrk-scrapers -e APIFY_TOKEN=your_token -- npx -y zrk-scrapers-mcp
```
**Claude Desktop, Cursor, Windsurf, VS Code** (`mcpServers` config)
```json
{
"mcpServers": {
"zrk-scrapers": {
"command": "npx",
"args": ["-y", "zrk-scrapers-mcp"],
"env": { "APIFY_TOKEN": "your_token" }
}
}
}
```
## Example prompts
- "Find remote product designer jobs at Linear, Figma and Notion posted this week."
- "Audit the mobile Core Web Vitals of the 20 main pages on example.com and list the worst offenders."
- "What Google Search ads is hubspot.com running in the UK right now? Give me the headlines."
- "Get the transcripts of @nasa's last 5 TikToks and summarize them."
- "Compare Booking.com vs the official site price for 3 hotels in Austin for Dec 12 to 14."
- "List software companies incorporated in Manchester in the last 7 days with their directors."
## Pricing and limits
Runs use your Apify account. Each actor is pay-per-result (for example $0.60 per 1,000 Telegram posts or $2 per 1,000 jobs); see each actor page for prices. Every run is capped at $1 by default (`maxChargeUsd` per call, or set `ZRK_MAX_CHARGE_USD` in env to change the default), and tools accept `maxItems` to limit the rows returned. Any other actor input field can be passed through `extraInput`.
Long runs return a `runId` after `waitSecs` (default 120); call `get_run_results` to collect them.
## License
MIT
TDQS
Scored across 10 tools
Each tool targets a clearly different data source and purpose: ATS jobs, Workday jobs, PageSpeed, UK companies, TikTok transcripts, Luma organizers, hotel prices, Google Ads, Telegram, and run results. The only nominal overlap is between the two job scrapers, but their descriptions make the ATS vs Workday boundary explicit. No two tools are likely to be confused in practice.
All names use snake_case and follow a verb_noun structure (scrape_, audit_, find_, get_). The verbs vary between scrape/get/find/audit, but they are semantically appropriate and do not break the overall pattern. This is predictable enough for an agent to infer behavior from names.
Ten tools is a well-scoped set for a multi-scraper server, with each tool representing a distinct Apify actor or shared run-management utility. There is no filler or redundant tool. The count is neither thin nor bloated.
The surface covers a broad range of scraper domains and includes get_run_results for asynchronous and paginated workflows. Minor gaps remain: there is no tool to list recent runs, cancel a run, or discover available actor options, which can create dead ends for run management. Core scraping workflows appear otherwise complete.