theprotocol.it Jobs Scraper

Scrape job offers from theprotocol.it, a Polish IT job board, into clean JSON, CSV or Excel: salaries (also normalised per month), required and optional skills, seniority, contract types, locations and company. No browser, no login: it reads the data theprotocol.it's own pages carry.

🔎 What does theprotocol.it Jobs Scraper do?

• Searches theprotocol.it by keywords, cities, categories, work type, experience level, contract type and "with salary only", or reads any listing or offer URL you paste.
• Returns one row per offer. With extractFullDetails on, each offer's own page is read too: full description (text and HTML), skills, benefits, address and company text.
• Runs on a schedule for daily new-job feeds, from the API, or from an AI agent.

⭐ Why this one

1️⃣ One schema across job portals. All our job actors share one output schema and one input schema. Build your pipeline once: our other job actors (NoFluffJobs, Just Join IT, RocketJobs.pl, Bulldogjob, Hollylist) take the same input and return the same fields, so adding a portal means changing the actor name, not your code.

• Low price, no start fee. $0.40 per 1,000 offers. The one other theprotocol.it scraper on the Store charges $1.00 per 1,000 and has a start fee. You pay only for rows you get.
• Salaries per month, plus the employer's own figures. Each pay range starts with a monthly entry; a pay per hour also keeps the employer's own figures as a second entry (isOriginal true). Each entry has min, max, currency, period, contractType and grossNet.
• Skills with a must-have flag. Every skill says whether the offer requires it or lists it as optional.
• Filters buyers ask for: search words, locations, categories, work type, experience level, contract type and "with salary only". Or paste any theprotocol.it listing or offer URL.
• Company data: with details on, each row carries the employer's active offers count and its own text about itself, once per company per run.
• Honest runs: a run summary and a health status. A search that fails writes an unbilled diagnostic row instead of silently returning nothing.
• Fast and cheap: plain page requests, no browser.

Data

This actor returns only what the site shows publicly at run time. We don't store or keep that data and are not its controller — you decide how you use it.

This is an independent tool, not affiliated with or endorsed by theprotocol.it; trademarks belong to their owners. It reads only public job pages: no login, no recruiter accounts. You are responsible for following theprotocol.it's terms and data-protection law (GDPR and others).

🤖 Use with AI agents (MCP)

AI assistants (Claude, ChatGPT, Cursor…) can run this actor through Apify's MCP server. Connect the server in your assistant, then ask, for example:

Use the theprotocol.it Jobs Scraper to find senior Python offers in Warsaw or fully remote that show a salary. Show the 10 best-paid, with company, salary, period and link.

⚙️ Input

Every job actor of ours takes the same inputs. Example: Python offers in Warsaw, B2B or employment, remote or hybrid, mid or senior, with full details:

{
  "searchQueries": ["python"],
  "locations": ["Warszawa"],
  "contractTypes": ["b2b", "employment"],
  "workType": ["remote", "hybrid"],
  "experience": ["mid", "senior"],
  "maxResults": 50,
  "extractFullDetails": true
}

Field	Default	Notes
maxResults	50	Offers in total. 0 = no limit (the run timeout still applies). theprotocol.it serves 50 offers per page
startUrls	–	theprotocol.it listing pages (https://theprotocol.it/praca, https://theprotocol.it/filtry/senior;p/zdalna;rw) or offer pages (https://theprotocol.it/praca/<slug>,oferta,<id>; /szczegoly/praca/... works too). A listing URL is read as that listing, newest first from page 1; an offer URL is always read from its own page. searchQueries is not applied (a warning says so). The filters below still apply: rows that do not match are dropped
searchQueries	–	Keywords, one search per line (a technology or a job title). Empty: every offer. theprotocol.it searches the whole offer text
locations	–	City names; accents and case do not matter (Krakow = Kraków = KRAKÓW). Any Polish city works; a city search also returns remote offers, as the site does. A city theprotocol.it does not know gives a warning and no search for it, never an unfiltered search. Empty: anywhere
categories	–	backend, frontend, fullstack, mobile, devops, data, ai-ml, testing, security, it-admin, support, embedded, gamedev, erp, architecture, project-management, product, ux-ui, business-analysis
contractTypes	–	employment, b2b, mandate, specific-task, internship
workType	–	remote, hybrid, onsite
experience	–	intern, junior, mid, senior, expert, lead, manager, executive
withSalaryOnly	false	Only offers that show a salary
extractFullDetails	false	Off: listing-level rows, cheaper and faster. On: each offer's own page (and its company, once per run) is read for the description, skills, benefits and company details. An offer URL is always read in full
maxConcurrency	1	Searches and detail pages processed in parallel
maxRequestsPerMinute	60	Request rate for the whole run
proxyConfiguration	off	Apify Proxy settings, useful when many detail pages are requested
maxRequestRetries	2	Retries after a failed request

theprotocol.it lists about 4,500 offers in total, 50 per page. For a smaller slice, split it with locations, categories or filters.

🧭 Input modes: what you give, what you get

One actor, six ways to use it. The mode follows from what you fill in.

Mode	You give	Rows	Offer pages read
🟢 1. Default search	nothing (empty input)	Newest offers, up to maxResults	Only with extractFullDetails on
🔤 2. Keyword search	searchQueries	One search per keyword	Only with extractFullDetails on
🎛️ 3. Filtered search	one or more filters, keywords optional	Offers matching the filters	Only with extractFullDetails on
🔗 4. Listing / search URL	startUrls with a listing URL	That listing, as the site shows it	Only with extractFullDetails on
📄 5. One offer URL	startUrls with an offer URL	That one offer	Always (a full row either way)
🧩 6. URLs + filters	startUrls plus filters	Only rows matching the filters	Offer URLs always; listed offers only with extractFullDetails on

Minimal input for each mode (real inputs from our test runs). Add "extractFullDetails": true to any of them for full rows.

🟢 1. Default search

{"maxResults": 3}

🔤 2. Keyword search

{"searchQueries": ["python"], "maxResults": 3}

🎛️ 3. Filtered search

{"searchQueries": ["python"], "locations": ["Warszawa"], "contractTypes": ["b2b", "employment"], "workType": ["remote", "hybrid"], "experience": ["senior"], "maxResults": 3}

🔗 4. Listing / search URL

{"startUrls": [{"url": "https://theprotocol.it/praca?kw=ai+devops"}], "maxResults": 3}

📄 5. One offer URL

{"startUrls": [{"url": "https://theprotocol.it/praca/senior-ai-devops-engineer-warszawa,oferta,01000000-326a-b7bb-0cf8-08df13fcc123?s=-10604388607&searchId=ce6f673a-5b8d-4044-9404-987b1b653b03"}]}

🧩 6. URLs + filters

{"startUrls": [{"url": "https://theprotocol.it/praca?kw=ai+devops"}, {"url": "https://theprotocol.it/praca/senior-ai-devops-engineer-warszawa,oferta,01000000-326a-b7bb-0cf8-08df13fcc123?s=-10604388607&searchId=ce6f673a-5b8d-4044-9404-987b1b653b03"}], "locations": ["Warszawa"], "contractTypes": ["b2b"], "workType": ["remote"], "experience": ["senior"], "withSalaryOnly": true, "maxResults": 3}

With details off, no offer page is read to check a filter: a row the listing alone cannot confirm is left out, and a warning in the run log says how many. Turn on extractFullDetails to check them.

📦 Output

A real row from a keyword search with details on (shortened; empty fields left out):

{
  "recordType": "job",
  "source": "theprotocol",
  "url": "https://theprotocol.it/szczegoly/praca/senior-python-developer-warszawa,oferta,01000000-af85-32bc-1c9b-08df25d4aa13",
  "title": "Senior Python Developer",
  "company": {
    "name": "ASTEK Polska",
    "profileUrl": "https://theprotocol.it/praca?kw=astek+polska",
    "activeOffersCount": 51
  },
  "locations": [{"city": "Warszawa", "region": "mazowieckie", "countryCode": "PL"}],
  "workType": [{"type": "hybrid"}],
  "categories": [{"value": "backend", "original": "Backend"}],
  "experience": {"levels": [{"value": "senior", "original": "senior", "minYears": 3}]},
  "contractTypes": [{"value": "employment", "original": "umowa o pracę"}, {"value": "b2b", "original": "kontrakt B2B"}],
  "workTime": "full-time",
  "salaries": [
    {"contractType": "employment", "currency": "PLN", "grossNet": "gross", "min": 16000, "max": 18000, "period": "month", "isOriginal": true},
    {"contractType": "b2b", "currency": "PLN", "grossNet": "net", "min": 20160, "max": 25200, "period": "month", "isOriginal": false},
    {"contractType": "b2b", "currency": "PLN", "grossNet": "net", "min": 120, "max": 150, "period": "hour", "isOriginal": true}
  ],
  "skills": [
    {"name": "Python", "required": true, "source": "portal"},
    {"name": "FastAPI", "required": true, "source": "portal"},
    {"name": "Django", "required": true, "source": "portal"}
  ],
  "benefits": ["Pakiet benefitów pracowniczych"],
  "languages": {"offer": "pl"},
  "description": {"text": "…", "html": "…"},
  "dates": {"posted": "2026-10-09", "expires": "2026-11-08", "scraped": "2026-10-09T07:18:46Z"},
  "extensions": {"isQuickApply": true, "openToUkrainians": false}
}

Missing data is null, never a guess.

Run status. After each run, open the run's key-value store (Storage tab) to see how it went:

• RUN_HEALTH: ok, or degraded with the reason (no rows, a failed request, a warning, or a row that broke a data rule).
• RUN_SUMMARY: the run's numbers: rows, requests, retries, failed requests, warnings and billed events.

💰 Pricing (Pay-Per-Event)

Event	Price	Trigger
Result row (job-scraped)	$0.0004 (= $0.40 / 1,000)	Per offer returned
Details row (job-details-scraped)	+$0.0015 (= +$1.50 / 1,000)	Per offer whose own page was read (extractFullDetails on, or an offer URL)

No start fee: a run costs only the rows it returns. Diagnostic rows are never billed. Example: 1,000 offers with details = $0.40 + $1.50 = $1.90.

🎯 Use cases

• Track new IT openings for a role or stack on a daily schedule.
• Feed a job board aggregator or a personal job-search dashboard.
• Compare pay by contract type, city and experience level, in the employers' own currencies.
• Market research: which skills and seniority levels are in demand.
• Lead lists of hiring companies.
• Run the same pipeline over several Polish job portals, since they share one schema.

❓ FAQ

How many offers can I get?

As many as maxResults says (0 = no limit). theprotocol.it serves 50 offers per page and lists about 4,500 offers in total; for a smaller slice, filter by city, category or experience.

Why is the salary empty for some offers?

Not every employer publishes pay. Then salaries is an empty list, never a guess. Turn on withSalaryOnly to get only offers with a salary.

What is the monthly salary?

Each pay range starts with a monthly item (period "month"). If the employer gives pay per hour, we compute it: hour × 168, rounded to a whole unit. The employer's own figures follow it with isOriginal true.

Can I combine it with other portals?

Yes. All our job actors share one output schema and one input schema, so rows from different portals fit one table. See the list below.

Can I run it on a schedule, through the API or from an AI agent?

Yes to all three: use Apify schedules, the Apify API and clients, or the MCP server (see "Use with AI agents" above).

What if my search finds nothing?

You pay nothing: no rows, no charge, and RUN_HEALTH says degraded (no rows). If the search failed (the site did not answer), you also get one unbilled diagnostic row (recordType "diagnostic") with the search and the reason in errorMessage.

🔗 Other Polish job scrapers (same schema)

• NoFluffJobs Jobs Scraper
• Just Join IT Jobs Scraper
• Bulldogjob Jobs Scraper
• RocketJobs.pl Jobs Scraper
• Hollylist Jobs Scraper (US entertainment jobs)

💬 Feedback

Found a missing field or a bug? Open an issue on the Issues tab.