JustJoin.it Tech Job Scraper (Poland)

Scrapes tech job offers from JustJoin.it — Poland's largest IT job board.

Public data only. No login, no cookies, no browser. No bot challenge of any kind — 6 of 6 TLS profiles returned clean 200s on every surface, including a bogus path (which answers a clean 404, not a challenge).

The one thing you need to know before using this

page= does nothing. JustJoin.it's search page is client-side infinite-scroll: the server always embeds the SAME freshest ~100 offers for a given (location, category, keyword) combination, whatever page number you ask for. There is no HTTP-only way to page past that — the client-side "load more" call was searched for across all 45 Next.js chunks referenced by the page and never found.

So this actor does not offer a page input at all. Coverage comes from diversifying the query: run several locations, or narrow with category / keyword, rather than expecting one broad search to page through everything. Each SEARCH_SUMMARY row reports likelyMoreThanCap: true when a query returned the full 100-offer cap, as a signal that narrowing it further would surface more.

What you get

Three record types share one dataset, told apart by recordType.

JOB — one row per offer

Search rows already carry a rich structured summary: title, company (name + logo), city, GPS coordinates, salary broken out per currency (PLN, EUR, USD, GBP, CHF) and contract type, experience level, workplace type (remote / hybrid / office), working time, required and nice-to-have skills, publish and expiry dates, and the apply URL.

With Fetch full job descriptions on (the default), each row also gets jobPostingJsonLd — JustJoin.it's Schema.org JobPosting record, verbatim, with the full plain-text description.

SEARCH_SUMMARY — one row per location searched

The filters actually sent, how many offers came back, and whether this query hit the 100-offer cap.

ERROR — one row per input that could not be processed

Every input maps to at least one row, so nothing disappears silently.

Input

Field	What it does
Locations	all-locations, or a city slug (warszawa, wroclaw, krakow…) — one search each
Category	one of 25 verified tech categories, or any
Keyword	free-text search box filter
Max offers per location	0 = unlimited, meaning up to the ~100-offer cap per query
Fetch full job descriptions	full HTML→text description via one extra request per offer
Min interval between requests	the speed control — 0.3s default

Example

{
  "locations": ["all-locations", "warszawa", "wroclaw"],
  "category": "python",
  "keyword": "senior",
  "maxItems": 100,
  "includeJobDetails": true
}

Notes on reliability

• No WAF. All surfaces answered 200 on chrome124, chrome120, safari17_0, chrome99_android and firefox133.
• Location and category are both honest. An unrecognised slug — alone or combined (warszawa/bogus-category) — answers a clean HTTP 404, never a silently widened baseline. A bogus keyword answers HTTP 200 with zero offers, also honest.
• Speed is controlled by request rate, not concurrency, same design as the other actors in this portfolio — maxConcurrency absorbs latency variance, minRequestInterval is the actual throttle.
• Residential proxy pinned to Poland is the cloud default.

Known limits

• ~100 offers per (location, category, keyword) combination, always the freshest ones — see "the one thing you need to know" above. There is no offset/cursor this actor can reach HTTP-only.
• The flight stream's own copy of the description is a chunked text-protocol reference this actor does not reassemble; the detail pass uses the page's JSON-LD block instead, which carries the same text unchunked. If you don't need the full description, list-only mode is one request per location regardless.