Blog

How do I get jobs straight from company career pages (Greenhouse, Lever, Ashby)?

2026-09-28

You should get jobs straight from company career pages because that's where the listing actually lives: the company controls it, the apply link goes to the real application, and when the role closes it disappears there first. For a lot of tech companies and startups, those career pages run on Greenhouse, Lever or Ashby, and Find your role first reads 8,519 of those boards directly.

Everything else you see on the internet is some kind of copy of those boards. Copies can be useful. They just drift.

What are Greenhouse, Lever and Ashby?

They're applicant tracking systems, the software companies use to post roles and manage applicants. When you click "Careers" on a startup's site and land on a clean list of openings, there's a good chance one of these three is behind it.

Each company's board is public. Anyone can open it in a browser, and each system also offers public endpoints for reading a company's open roles. That's what makes a direct feed possible.

Where do other job listings come from?

Mostly from those same boards, one or two steps removed. A job aggregator reads a board, or reads another site that read the board, and republishes the role. Some companies also post separately to LinkedIn or niche job sites.

Nothing wrong with that on its own. The issues come from the distance:

• Timing. A copy can only be as fresh as the last time it was collected, and then it still has to be indexed and shown to you.
• Duplicates. The same role can appear once per site that copied it, sometimes with slightly different titles.
• Closed roles. When a company closes a role on its board, every copy has to notice. Some take a while.
• Apply links. A copy sometimes sends you through its own page, or a redirect, before you reach the real form.

To be fair, aggregators also do real work. They cover companies on every applicant system, including many that aren't on these three, and they make search easy for people browsing by hand.

Why does this matter more for an AI agent?

A person can spot a stale listing and move on. An agent is more literal. It'll happily include a closed role in your shortlist, or read the same job twice under two titles, or follow a redirect chain to a login wall. Every one of those wastes a tool call and a bit of your trust in the results.

Reading from the source cuts most of that out. One role, one record, one real apply link, and a closed status that comes from the company's own board.

What Find your role first reads, and how often

The feed covers 8,519 company boards across Greenhouse, Lever and Ashby, which adds up to about 250,000 open roles. No job board middleman, no scraped aggregators.

Every board is checked every two hours. The checks are staggered across that window, so new roles show up continuously instead of in one big batch. On a normal day that's dozens of new roles every hour.

Each role carries two dates: when the company published it, and when the feed first discovered it. That turns out to be useful, since a board's published date and the moment a role actually went live don't always match.

What you get back

For each role: title, company, location, dates, the original posting link on the company's board, and the apply link.

The full job description stays on the company's page. Your agent opens the original posting with its own web or browser tool when it wants the details. That way it always reads the version the company is maintaining right now.

Could you just scrape the boards yourself?

Yes, and honestly, if you only care about twenty companies, you probably should. Point a script at their boards and you're done.

It gets harder at scale. You need the list of boards in the first place, and that list keeps changing as companies start hiring, stop hiring or move between systems. You need to check them on a schedule, notice closed roles, dedupe, handle the differences between the three systems, and store the history so "new since yesterday" means something. Then you need to put all of it behind something your agent can call.

That's the part the feed does. You pay $9 a month and skip the maintenance.

The honest limit

This only covers companies whose career pages run on Greenhouse, Lever or Ashby. Plenty of large companies use other systems, and those won't show up here. If your target list is mostly big enterprises, you'll want another source alongside this one.

And applying stays with you. The feed finds the role and gives you the link.

How to try it

Sign in with Google at findyourrolefirst.click/account, subscribe, create an agent key, and paste one line.

Claude Code:

claude mcp add --transport http --scope user jobfeed https://findyourrolefirst.click/mcp --header 'Authorization: Bearer YOUR_API_KEY'

Codex CLI:

export JOBFEED_API_KEY=YOUR_API_KEY
codex mcp add jobfeed --url https://findyourrolefirst.click/mcp --bearer-token-env-var JOBFEED_API_KEY

Other HTTP MCP clients: use https://findyourrolefirst.click/mcp with an Authorization: Bearer YOUR_API_KEY header.

Questions

Where does Find your role first get its jobs?

Directly from company career boards hosted on Greenhouse, Lever and Ashby. It reads 8,519 of them and doesn't use aggregators.

How many companies does it cover?

8,519 company boards, with about 250,000 open roles between them.

Does it include companies that don't use Greenhouse, Lever or Ashby?

No. It only reads those three systems, so companies on other applicant tracking systems aren't included.

How often are the career pages checked?

Every board is checked every two hours, with checks spread out so new roles arrive all day.

Is the apply link the company's own?

Yes. Each record has the original posting link and the apply link from the company's board.