Skip to main content

Customization guide

Use this guide when you want to change what jobpull keeps, how it classifies jobs, and where it runs.

Start with a pack​

Starter packs are the fastest way to launch a board with sane defaults.

npm run cli -- packs list
npm run cli -- packs show product
npm run cli -- packs apply engineering

Available packs:

  • product
  • engineering
  • design
  • gtm

Pack application updates role targeting, focus categories, tag keywords, and managed starter sources without deleting custom sources.

Tune the role filters​

The board keeps jobs based on the patterns in worker/src/config.ts.

Common changes:

  • adjust titleIncludePatterns to match the roles you want
  • adjust titleExcludePatterns to remove adjacent roles
  • change remoteOnly if you want onsite or hybrid roles
  • edit focusCategories to control the filters shown in the UI
  • update tagKeywords to drive auto-tags and digest grouping

Add sources​

Prefer the CLI:

npm run cli -- sources

The built-in read-only providers are:

  • Greenhouse
  • Lever
  • Ashby
  • Recruitee
  • Personio XML
  • Generic JSON

See Source Feeds for the exact config shapes and validation rules.

Brand the board​

Update the board name and tagline through the setup flow or the configuration CLI, then use the docs and landing page copy as a reference for the public-facing language.

Manage the schedule​

The default crawl cadence is daily. For the Cloudflare path, update the cron in worker/wrangler.toml. For Docker, change CRON_SCHEDULE in .env.docker.

Handle stale jobs​

jobpull soft-hides stale jobs instead of deleting them. Re-seen jobs are reactivated on the next crawl.

When to use Docker​

Use Docker Compose if you want:

  • one portable app container
  • same-origin API and frontend
  • SQLite on a VPS or local machine
  • a deployment path that is not tied to Cloudflare