LinkedIn Job Scraper (PPS)

Why it exists

A recruitment business needs to know who is hiring, where, and for what — every day. Manual LinkedIn searching cannot keep up, and paid data vendors are expensive.

What it does

• Scrapes public LinkedIn job search pages with Playwright, in a fresh logged-out browser context each time.
• Enriches each listing with the job detail page and company information.
• Cleans and de-duplicates across runs so the same job is never counted twice.
• Filters by title, location, and work type.
• Exports to CSV and shows results in a small web dashboard.

Engineering notes

• Strict operating rules are documented in the repo: never log in, never bypass a block, never make a live request without approval.
• Every run writes a status report. A silent-truncation bug (runs that finished “clean” but skipped detail pages) was found through these reports and fixed with explicit accounting of attempted versus collected pages.
• The scraper feeds two other tools: the ERS CRM and the Agency Signal analysis.