Every profile claims the same stack. hyreflow reads the code people have published and the technology companies really run, so your shortlist and your account list rest on evidence.
Claude
hyreflow connectedUse hyreflow and source Rust engineers in Germany on proof of work. Show me the code behind each name, and do not limit it to people who are active in open source.
A developer role, or the one technology your desk is built around.
hyreflow reads public GitHub work to see who has built with it, merges that with a wider market search so nobody is missed for being quiet in open source, and finds the companies that run the technology or are hiring around it, with the engineering leader at each.
You see the evidence behind every name and every company. By default it stops there before spending on contact details.
Each one answers a question a job title cannot.
Public code
Who has built with a technology, shown in work you can open and read.
Public work only. Much of any engineer's code is private, so this is a floor, and the search is merged with a wider market pull.
Technology detections and funding
Which companies in your patch run the technology you recruit for, whether or not they are hiring today.
Detects what leaves a public trace. Internal tools often show up in job adverts first, which is why the plays read both. Website-level stack data is available with your own BuiltWith account.
Live job adverts
Who is hiring around your technology right now, and what else is in their stack.
Sees advertised roles. A company hiring through referrals alone will not appear here.
A signal is a reason to look, not a guarantee. Each one links to the play that works it.
The strongest account on the list: the need is real and already budgeted.
A standing account. Worth knowing the engineering leader before the role exists.
A short window to reach the engineers affected, and often an opening with the employer that remains.
Engineering roles tend to follow. The play marks the likely roles as inference and builds a small pool for each.
Claudehyreflow connectedUse hyreflow and build me the list of every company in the Netherlands running Kubernetes at scale, with the engineering leader at each.
Name the technology your desk is built around. hyreflow returns the companies in your geography that run it, each with the source and dates behind the detection, a tier against your ideal client, and the engineering leader with a checked work email.
Every play is a prompt you can copy, with the tools it calls and what it costs to run.
On what they have built
On what they run
GitHub, technology detections and job-advert data are part of hyreflow. Your ATS and your sequencer plug in through the account you already have.
The general method behind the shortlist: scored on evidence, with the reason written down for every name.
See sourcingWhat Claude can and cannot do on its own, how to give it candidate data, and the prompt that builds a shortlist.
Niklas HuetzenRead the guideNo install. Add hyreflow to Claude or ChatGPT as a connector and run your first play.
and any MCP clientConnect hyreflowSomething else? Bring it to Open Office, the weekly open call.
It searches public repositories and their contributors for the technology in your brief, reads each person's public work to see whether the skill claim is backed by code, and scores them against the role. That list is merged with a wider market search, so you are not limited to people who are active in open source.
No. Much of any engineer's work is private, so GitHub is treated as evidence where it exists, not as the whole pool. The play runs a broad people search alongside it and marks which candidates have public work behind them.
Sometimes a developer publishes one, and many do not. Contact details come from hyreflow's data providers instead: a personal email where one exists, and the LinkedIn profile where it does not. Candidates are never approached on a work email.
From two directions. Technology detections show what a company runs, with the source and dates behind each one. Job adverts show what a company names when it hires. The account list play reads both, removes duplicates, and shows the source and dates behind every detection.
Yes, from the company's public GitHub organisation: repositories and languages, who contributes to each, who the maintainers are, and which of them look like staff. It is a floor, because much of any team is not visible in public.
The general sourcing plays cover product, data, design and go-to-market roles in technology companies. The GitHub play is specific to roles where public code is evidence.
Connect hyreflow to Claude or ChatGPT, attach the spec, and run the GitHub sourcing play. Every name comes with the work behind it.
