Overview

This folder contains a complete, standalone Apify Actor for rich career-site and applicant-tracking-system job data. Its project layout follows the workspace’s jobright-scraper: executable code is under src/, Apify definitions are under .actor/, the container is defined by Dockerfile, and dataset-validator.js checks every locally generated record. The Actor does not import code from any sibling scraper, so it can be pushed, built, scheduled, and versioned on its own.

Live mode calls the official Fantastic.jobs /v1/active-ats endpoint. It requests descriptions, public company enrichment, occupation classifications, and structured geocoded locations, then normalizes the response into a stable schema. Live runs require fantasticJobsApiKey or the FANTASTIC_JOBS_API_KEY environment variable. The secret is used only in the Authorization header; it is never printed, stored in the dataset, or written to diagnostics.

Rich output

The normalized record includes stable IDs, title, company, all locations, salary range and period, employment type, workplace mode, dates, full description text and HTML, responsibilities, requirements, benefits, skills, keywords, taxonomies, education, language, visa sponsorship, application URLs, recruiter details, hiring-manager fields, LinkedIn organization enrichment, company size and funding, occupational classifications, and structured locations. Every upstream field remains available under rawData.

No source value is invented. Null and undefined values are removed recursively from rawData. Normalized optional fields use empty strings, arrays, or objects, so the dataset contains no JSON null values while still distinguishing “not published” from factual data. extractionStatus, dataMode, and sampleProvenance make the origin and completeness of each record explicit.

Inputs

The complete machine-readable contract is in .actor/input_schema.json. Useful inputs are keyword, location, organization, timeRange, descriptionFormat, includeCompanyDetails, includeBetaFields, and advancedFilters. maxItems defaults to ten and the runtime enforces a minimum request of ten. Advanced filters accept documented primitive Fantastic.jobs query parameters; input is never evaluated as code.

Local development

Install dependencies and run an explicit official-sample contract test:

npm install
apify run --purge --input '{"sampleMode":true,"maxItems":10}'
npm run validate:data

Then run live with a secret key:

apify run --purge --input '{"fantasticJobsApiKey":"<secret>","keyword":"software engineer","timeRange":"7d","maxItems":10}'
npm run validate:data

Sample mode reads Fantastic.jobs’ published ATS sample and labels every record official_sample. It verifies normalization, schema, storage, and validation only; it must not be presented as a live search result. Live mode never falls back silently to sample data.

Run status and production

The default dataset contains records. RUN_SUMMARY reports status, requested and returned counts, provenance mode, timing, and null-validation count. SOURCE_METADATA records the provider and endpoint. A failure creates a sanitized ERROR_DIAGNOSTIC and never creates fake jobs or misclassifies an access or credential error as no_results.

Deploy this folder independently with apify push. Configure secrets in Apify’s encrypted input or environment settings, not in source files. Respect upstream API plan limits, terms, privacy requirements, and retention rules when scheduling or redistributing the public job data.

Source identity

• Standalone Actor: career-site-job-listing-api
• Original catalog reference ID: s3dtSTZSZWFtAVLn5
• Provider: Fantastic.jobs active ATS API
• Record type: rich job
• Default output request: 10

Live source update (2026-10-21)

The default dataSource: "direct" reads current public Greenhouse, Ashby, and Lever board APIs from startUrls. No vendor API key is needed. Default URLs are prefilled in the input form. Full source descriptions and provenance are returned; no sample or synthetic fallback is allowed. Empty valid searches return zero records; unavailable sources fail instead of pretending success.

Use dataSource: "fantastic" with your fantasticJobsApiKey for the original paid feed, vendor time windows, and enrichment. Direct mode does not provide Fantastic.jobs enrichment or historical time-window filtering. The older vendor-specific reference below applies only to Fantastic mode.