About the JobSpecter job reader

JobSpecter helps a small group of job seekers find openings that fit them. It reads public job listings from company career sites, the same listings anyone can open in a browser, and nothing else. It never applies to jobs, never signs in, and never fills in forms on your site.

How it identifies itself

Every request carries this User-Agent:

Mozilla/5.0 (compatible; JobSpecter/2.5.0; +https://jobspecter.com/bot)

Requests come from one fixed address. We don't rotate addresses, use proxy pools or disguise the reader.

How politely it reads

• One request at a time per site, at least 2.0 seconds apart by default (job-board APIs designed for this, such as Greenhouse or Lever, at least 1.0 seconds).
• A daily request budget per site (by default 2500 requests).
• Most companies are read once every one to seven days, depending on whether anyone follows them, at a fixed time spread across the day. Unchanged lists are detected with ETag / Last-Modified where offered.
• It follows robots.txt for normal web pages (agent name JobSpecter). Public job APIs meant for programmatic reading are read as documented.
• If your site answers 429, 403 or a challenge page, the reader stops and waits: 6, then 24, then 72 hours. It honours Retry-After.

Ask us to stop or slow down

Add a Disallow rule for User-agent: JobSpecter in your robots.txt, or answer with 429, and we will back off automatically.