Greenhouse Jobs Scraper

Job postings from public Greenhouse job boards (boards.greenhouse.io/{board}), read from the public Job Board API that the boards use. One item per job, with the description as clean text and HTML, and recruiter contact details removed.

Independent tool. Not affiliated with, endorsed or sponsored by Greenhouse Software, Inc. or any employer whose board it reads. Company names and trademarks belong to their owners.

What it does

• Takes one or more board tokens or board URLs: stripe, https://boards.greenhouse.io/stripe or https://job-boards.greenhouse.io/stripe. Any other host (for example a company site with a ?gh_jid= link) gets an error item with no charge.
• Makes one request per board: GET https://boards-api.greenhouse.io/v1/boards/{board}/jobs, with ?content=true only when includeDescription is on. No authentication is used.
• Filters run on the data already downloaded. There are no per-job requests.
• Never logs in and never calls application endpoints. No candidate data is read or stored.

Input

{
  "boards": ["stripe"],
  "searchText": "engineer",
  "departmentContains": "Engineering",
  "locationContains": "Dublin",
  "postedWithinDays": 30,
  "maxItems": 20,
  "includeDescription": true
}

Field	Type	Default	Notes
boards	string[]	required	Board tokens or board URLs. Max 50 per run. Duplicates count once.
searchText	string	—	Case- and accent-insensitive. Matches the title and, when includeDescription is on, the description.
locationContains	string	—	Case- and accent-insensitive, on the job location.
departmentContains	string	—	Case- and accent-insensitive, on the department names.
postedWithinDays	integer ≥ 0	—	UTC calendar days from the first published date (the updated date if that is missing). 0 keeps jobs published today.
maxItems	integer	50	Items per run. Hard cap 1000, enforced in code.
includeDescription	boolean	true	Adds descriptionText and descriptionHtml. See "Large boards" below.

Output (one item per job)

{
  "company": "Figma",
  "title": "Account Executive, Enterprise",
  "jobId": "5426468004",
  "url": "https://boards.greenhouse.io/figma/jobs/5426468004?gh_jid=5426468004",
  "location": "San Francisco, CA • New York, NY • United States",
  "departments": ["Sales"],
  "offices": ["US"],
  "postedDate": "2025-01-28T23:57:29.000Z",
  "updatedAt": "2026-07-22T09:37:08.000Z",
  "descriptionText": "Figma is growing our team of passionate creatives ...",
  "descriptionHtml": "<div class=\"content-intro\"><p>Figma is growing ...",
  "source_url": "https://boards.greenhouse.io/figma",
  "retrieved_at": "2026-10-23T16:16:02.486Z",
  "source_name": "Greenhouse",
  "status": "ok",
  "error_message": null,
  "warnings": ["contato_removido"]
}

• status is ok (a job, charged) or erro (a board or URL that could not be read, free). error_message says why.
• postedDate is Greenhouse's first_published; updatedAt is updated_at. Both are UTC.
• descriptionHtml is null when includeDescription is off.
• Warnings: contato_removido (a contact was removed), descricao_indisponivel (the job has no description in the response, so it is returned without one and not charged), teto_margem:<board>, maxItems_limitado:<n>.
• The key-value record RUN_SUMMARY has the run status (ok, sem_resultado, parcial, fonte_indisponivel, erro_url, erro_input), completo, the stop reason motivo (maxItems, teto_cobranca, prazo_interno, teto_margem), per-board HTTP status and counts, the charged events and a margem block.

Pricing (pay per event)

Event	When	Proposed price
inicio-run	Once per run, after the input is valid and at least one board is valid	US$0.005
job	Each job with status ok. A job whose description was unavailable (warning descricao_indisponivel) is returned and not charged	US$0.001

Invalid input and runs where no board is valid are not charged. Error items are free. The maximum charge per run (ACTOR_MAX_TOTAL_CHARGE_USD) is respected.

Limits

• Memory 256 MB, timeout 300 s. The Actor stops on its own at 270 s (or 20 s before the platform timeout, if that comes first) and keeps what it already saved.
• Requests run one at a time, at least 1 s apart. Up to 3 tries on HTTP 429 and 5xx, with Retry-After honored up to 10 s.
• A source failure never fails the run. A board that is down, returns 404 or returns a non-JSON page gets an erro item, and the run ends SUCCEEDED.
• Download budget and large boards. With descriptions on, a board comes in one response that includes every description. Before each request the Actor reserves the download its revenue can pay for: with only the start charge (US$0.005) that is about 15.8 MB, and each charged job adds about 4 MB. A board bigger than that is cut off mid-download and stops the run with motivo: teto_margem and nothing charged. Measured on 23/10/2026 with content=true: Stripe 5.6 MB (722 jobs), Figma 1.6 MB (140 jobs), Vercel 0.9 MB (95 jobs), Mercury 0.7 MB (75 jobs). Without descriptions the same boards are much smaller (Stripe: about 450 KB).
• Cost guard. Before each request the Actor checks that the estimated cost (compute time since the start, plus a 20 s tail for the request in flight, plus download bytes at US$0.20/GB) stays under 80% of the revenue charged so far, minus US$0.0005. When it would not, the run stops with motivo: teto_margem and status: parcial.
• Search is local. Greenhouse has no server-side search, so searchText, locationContains and departmentContains are applied after the board is downloaded.

Polite crawling

• Public, unauthenticated GET requests only. One request at a time, 1 s apart.
• Check each board's terms of use before large runs. You are responsible for how you use the data.

Data use and privacy

Job postings are company data. The only personal data that can appear is a recruiter's contact inside the free text. The Actor removes:

• every e-mail address ([email removido]), including company inboxes such as an accommodations address,
• Brazilian and US-format phone numbers ([telefone removido]),
• lines such as Recruiter: Name, Recrutador(a): Nome or Talent Acquisition: Name, in the text and in the HTML,

and adds the contato_removido warning. No candidate data is read or stored.

The filter is pattern based. Check the output before you use it for mailings or a CRM, and make sure your use meets the rules that apply to you (for example LGPD or GDPR).

Known limits

• Phone detection covers Brazilian and US formats. Other formats may pass through.
• A board with no open jobs returns sem_resultado with zero items.
• The download allowance applies to the whole board response (see "Large boards with descriptions").
• Machine values (status ok/erro, the warning codes, the RUN_SUMMARY keys and the motivo values) are in Portuguese, as in the sister Actors. Field names and prose are in English.

Run locally

node --test                    # unit tests, no network
node src/main.js               # local run; input in storage/key_value_stores/default/INPUT.json
node scripts/evidencia.mjs     # regenerates prova/run-local and prova/conferencia.md

---

Greenhouse Jobs Scraper (português)

Vagas publicadas em boards do Greenhouse públicos, lidas pela API pública que os próprios boards usam. Um item por vaga, com a descrição em texto limpo e em HTML, sem contato pessoal de recrutador.

Entrada: tokens de board ou URLs de boards.greenhouse.io e job-boards.greenhouse.io (boards). Outros hosts, como link ?gh_jid= de site da empresa, viram item erro, sem cobrança. Saída: um item por vaga (company, title, jobId, url, location, departments, postedDate, descriptionText, descriptionHtml, status, warnings etc.). Erros de fonte não derrubam o run: ele termina com sucesso e registra um item erro.

Preços: inicio-run US$0,005 por run com board válido; job US$0,001 por vaga com status: ok e descrição. Erros e vagas sem descrição disponível não são cobrados.

Limites: até 50 boards por run; maxItems padrão 50, teto de 1000 no código; 256 MB; prazo interno de 270 s; requisições uma a uma, com 1 s entre elas. Um board com descrição maior que a folga de download (cerca de 15,8 MB no início do run) para com teto_margem e não cobra nada. O Stripe, com descrição, tem 5,6 MB e cabe.

Privacidade (LGPD): a vaga é dado da empresa. E-mails, telefones (BR e EUA) e linhas do tipo Recrutador(a): Nome saem da descrição, com o aviso contato_removido. Nenhum dado de candidato é lido ou guardado. O uso dos dados é responsabilidade de quem usa.

Related Actors

More Actors from the same publisher:

• Lever Jobs Scraper API: Job postings from public Lever job boards, with clean descriptions.
• Ashby Jobs Scraper API: Job postings from public Ashby job boards, with department, employment type and pay summary.
• Workday Jobs Scraper API: Job postings from public Workday career sites (myworkdayjobs.com), with clean descriptions.
• Brazil Jobs Scraper: Greenhouse & Lever (Vagas BR): Job openings at Brazilian companies from public Greenhouse and Lever job boards, with normalized city, state and work model.

Browse the full catalog: kman-br.github.io (Português).