Himalayas Jobs Search Scraper

This Apify Actor extracts rich remote-job records from the official public Himalayas API. It runs locally or on Apify Cloud using bounded native HTTPS, with optional standard proxy routing when direct access is unavailable. It does not use fingerprint spoofing, stealth, or CAPTCHA/WAF bypasses.

Modes

• search: query the official API and emit complete current records.
• searchMultiple: interleave bounded results from queries or searchQueries.
• single: resolve one official company/job URL through the API.
• multiple: resolve jobUrls or urls through the API.
• startUrls: accept official job URLs or /jobs search URLs; search URLs are translated to the public API query.

The public HTML jobs route is Cloudflare-gated in the local environment. Rows therefore state detailFetched: false and detailVerified: false; they are verified from the official API record and never pretend to have a detail-page fetch. An API applicationLink equal to the job page is retained as sourceApplicationLink, while applicationUrl is omitted unless a distinct source URL is published.

Local run

npm ci --ignore-scripts
npm test
npm run check
apify validate-schema
apify run --purge --input-file INPUT.json
npm run validate

Dataset rows include truthful executionEnvironment and proxyUsed fields. Run health, diagnostics, skips, metadata, and API receipts are written to the default key-value store as RUN_SUMMARY, RUN_DIAGNOSTICS, RUN_SKIPS, RUN_HEALTH, RUN_METADATA, and RUN_REQUESTS. status reports only the Apify platform outcome (SUCCEEDED or FAILED); resultStatus reports data completeness (COMPLETE, LIMITED, SKIPPED, DEFERRED, NO_DATA, or FAILED). A handled source barrier is SUCCEEDED/SKIPPED; uncertain source access is SUCCEEDED/DEFERRED; a clean empty API result is SUCCEEDED/NO_DATA; and a fatal zero-row error is FAILED/FAILED. If some real rows were persisted before a later non-block failure, they remain in the Dataset and the run is SUCCEEDED/LIMITED. Diagnostics remain in key-value storage, never as Dataset rows.

Search input

{
  "mode": "search",
  "query": "developer",
  "location": "Remote",
  "maxItems": 3,
  "maxCandidates": 12,
  "maxPages": 1,
  "maxRequests": 12,
  "requestTimeoutSecs": 20,
  "maxRetries": 2,
  "proxyConfiguration": { "useApifyProxy": false }
}

Every output row is buffered until its official GUID, title, company, and non-empty rich description are present. Missing optional salary, category, benefit, or application fields are omitted rather than replaced with null, blank, or guessed values. Partial runs preserve valid rows and report diagnostics; fatal zero-row errors use the Actor failure lifecycle. HTTP 401/403/429/451 or visible source-authored access challenges stop requests without retry and are recorded as SKIPPED; generic transport and HTTP errors are not mislabeled as blocks. Keep request and item limits bounded, and collect only public job data in accordance with applicable terms and laws.