CareerBuilder Public Jobs Search Scraper

Collect complete, detail-verified CareerBuilder postings from public search pages. Records are emitted only when the source exposes a readable description and more than 20 meaningful facts tied to the job.

Current access status

CareerBuilder's ordinary public access path returned HTTP 403 during the prior source audit. The board and a representative Broadridge detail were visible in Chrome, but the Actor's direct request to the same public detail URL was denied. Per the repository access policy, do not run the Actor against CareerBuilder again during this blocked period, retry the request, change routes or identity, use cookies or a proxy, or try to solve a challenge. The Actor package is retained with single-attempt direct access and immediate SKIPPED handling so its behavior is ready for a future pass only after the public access state changes. That handled skip is a successful platform run (status: SUCCEEDED) with resultStatus: SKIPPED; it does not mean data was extracted.

Data fields

When the public response exposes them, job rows include the source identifier, title, employer and public company facts, location components, employment and workplace labels, dates, readable description and sections, responsibilities, qualifications, skills, benefits, salary and unit, application URL, detail provenance, and source-fact count. Salary and other optional facts are omitted when they are not tied to the posting.

The dataset schema documents source facts and request provenance. Each emitted job must contain more than 20 populated meaningful source facts. Diagnostics are stored separately from the dataset under RUN_DIAGNOSTICS; request evidence is stored under REQUEST_RECEIPTS and RUN_SUMMARY.

Input contract

The input schema accepts keyword, location, up to five exact CareerBuilder /jobs or /job-details/ URLs, maxItems (1–10), timeoutMs (5,000–60,000), and delayMillis (0–3,000). Unsupported fields, fixture switches, arbitrary URLs, custom headers, cookies, proxies, and retry controls are rejected.

Example input for a future run after the public route is no longer blocked:

{
  "keyword": "software engineer",
  "location": "New York, NY",
  "maxItems": 2,
  "timeoutMs": 30000,
  "delayMillis": 750
}

Access policy and validation

Requests are direct, sequential, and single-attempt. HTTP 401, 403, 429, 451 or visible source-authored denial/challenge content stops requests immediately, discards buffered job records, and records status: SUCCEEDED / resultStatus: SKIPPED with request receipts and diagnostics in the key-value store. Redirects, other HTTP errors, oversized responses, and transport failures are SUCCEEDED / DEFERRED, not source blocks. Complete data is SUCCEEDED / COMPLETE; stored rows followed by a non-block failure are SUCCEEDED / LIMITED; fatal code, input, artifact, or dataset failures before any stored row use FAILED / FAILED and the Actor failure lifecycle. This Actor has no authoritative empty-results detector: noDataConfirmed remains false and unrecognized empty/job-shaped responses are SUCCEEDED / DEFERRED, never NO_DATA. Diagnostics are never dataset items. The Actor does not use a proxy, cookies, custom headers, fingerprint changes, or alternate routes.

Offline checks:

npm test
npm run check
apify validate-schema

Dataset reconciliation for an actual authorized local run:

npm run validate -- <fresh-actor-relative-storage-directory>

No CareerBuilder dataset was qualified in the current pass because the public route is blocked. This Actor is not affiliated with CareerBuilder. Use source data only in accordance with applicable site terms and law.