demo-data

Static, self-hosted test fixtures for QA-ing document/web extraction and browsing tools. See manifest.json for which file satisfies which test case, and the repo README for how these are generated.

All company and competitor names below are fictional, modelled on a real customer's industry without naming them.

Multi-page careers site — simple case

Plain static HTML, no JS, no query-string pagination — deliberately "simple case, not SPA".

Careers site — SPA (hard case)

Genuinely client-rendered: content only exists after JS runs, and search/pagination/job-detail are all click-driven with no URL change at all (no ?q=, no ?page=2, no separate page load). Same 6 roles as the simple case above, for direct comparison.

Shared source content

Used by the role/content/market fixtures below — one company's value proposition and employee stories.

Role fixtures

Single role + single candidate pool scenario.

Content coverage fixtures

Role list vs. testimonial coverage scenario, deliberately uneven coverage.

Market fixtures

Home-market vs. target-market adaptation scenario — two target-market research reports.

Sales / Commercial — Fenwick Retail Group

Software Engineering — Cascadia Cloud Systems

Hospitality Operations — Northbridge Hotels & Resorts

Edge cases

Live-search-only acceptance criteria

No fixture needed — see notCovered in manifest.json for suggested real search targets (employer comparisons, talent-pool research, etc.).