# Changelog of Greenhouse Jobs Scraper – Job Listings by Company Board (`rowcraft/greenhouse-jobs-scraper`) Actor

- **URL**: https://apify.com/rowcraft/greenhouse-jobs-scraper/changelog.md
- **Full Actor documentation**: https://apify.com/rowcraft/greenhouse-jobs-scraper.md

## Changelog

### 1.1.9 (2026-10-17)

- README: the list of other Rowcraft actors now includes the five website and extension tools published on 16.10.2026.
- Shared library: URL-like fields named link or guid (feed items) are no longer flagged as phone numbers; retry log line now reads "retry n/m".
- Shared library: an HTML page where JSON was expected (maintenance or block page, seen on 17.10.2026 at the Polish KRS API and at Workday) is retried and then reported with the page text instead of crashing the run on a parse error.

### 1.1.8 (2026-10-15)

- README: the list of other Rowcraft actors now includes the five SEO and domain tools published on 15.10.2026.

### 1.1.7 (2026-10-13)

- README: the list of other Rowcraft actors now covers the whole family, and every ready-made example is linked from the Use cases section.
- Memory: the short-lived part of the Node heap is capped, so large runs on small memory no longer end without a message; the run summary reports the memory peak (memoryPeakMb).

### 1.1.6 (2026-10-13)

- Long runs over many companies no longer run out of memory. Node reserved up to 192 MB of the run's 256 MB for short-lived objects; that reserve is now kept small, so a run over hundreds of boards stays well under the limit instead of climbing past it and ending without a message. The run summary shows the peak memory (`memoryPeakMb`) and the log reports it every 20 companies.
- A board that is still above 15 MB without the ad texts (Pulse Healthcare: 20.8 MB) is read at the default 256 MB too; the limit of that read grows with the memory you give the run.

### 1.1.5 (2026-10-12)

- Very large Greenhouse career pages no longer stop the run. A board with thousands of jobs (for example Pulse Healthcare: 2,728 jobs, 39 MB with the ad texts) used to fill the run's 256 MB of memory and end the run without a message, losing the coverage report. Such a board is now read without the texts in one request, and each delivered job's text is read with its own request.
- An answer too large for the run's memory is now reported for that company (in the log and in `COVERAGE`) and the run goes on with the next one. The limit grows with the memory you give the run.

### 1.1.4 (2026-10-12)

- A job link whose job has closed is now named: the company's coverage line says the linked job is no longer open and gives its id in `missingJobLinks` (also in the run summary). Before, it read "no open jobs on the board".
- Location lines such as "United States - Remote" now keep the country (`countryCode` US) instead of reading "United States -" as a city.

### 1.1.3 (2026-10-11)

- Workplace type (`workplaceType`): employers' own labels now map to On-site, Hybrid or Remote (for example Workday's "Office - Flexible" becomes Hybrid and "Remote Home-Based" becomes Remote); for Workday jobs the employer's wording stays in `metadata.workplaceLabel`.
- README: screenshots of the input form and of the results table, taken from real runs.

### 1.1.2 (2026-10-11)

- If a career page gives no company name, the name from the ready-made company lists is used.

### 1.1.1 (2026-10-11)

- The field `stateKey` is now `stateName` (the memory used by **Only new rows**): the Apify quality check treats a field named \*Key as a secret and asked for it to be encrypted, which would hide the name you chose. Runs that still send `stateKey` keep working.
- The webhook address in `notifyUrl` is stored encrypted (secret input field): webhook URLs are credentials.

### 1.1.0 (2026-10-10)

- Ready-made lists: every company's name now comes from its own career page (Ashby, Lever, Personio) or job API, no longer from the seed directory; 20 list companies used to show another company's name.
- Industry lists keep only companies verified as the seed company (same website, or the board name is the company's own web address).
- A ready-made list is offered only when it holds at least 5 companies of this actor's system(s).

### 1.0.0 (2026-10-10)

- First release: Greenhouse career pages read through their public job APIs; one row shape for every system (title, locations with country code, department, team, employment and workplace type, salary when published, texts with contact details removed).
- Ready-made company lists verified on the public APIs, keyword / location / department / type / date / salary filters, per-company coverage report.
- Standard Rowcraft automation: only new rows since the last run (`onlyNew`, `stateKey`), end-of-run notification (`notifyUrl`), charge cap respected, one request per second.