Agentic SQA is generally available. Connect a repository and it reviews your next pull request.
Agentic SQA is evidence-first verification for systems code. It investigates pull requests in C codebases, finds likely defects and reports each one with the evidence to act on it: the suspected failure mode, the exact scope and the next step, split into what to do first and what deserves deeper inspection. It ships through the IR Labs GitHub App, Foundry Verify, and findings land as comments on the pull request with a check-run alongside.
Why we built it
It keeps getting cheaper to generate code and cheaper to generate opinions about code. Every tool can leave a plausible remark on a pull request now, and in systems code plausible is not the bar. A maintainer of a driver, a kernel module or a firmware tree needs to know which risks are real before they merge, and that is a verification problem, not a commentary problem. We wrote about the design in more depth in The evidence layer.
Two decisions define the product. Every finding carries its reasoning and its scope, so you can reject it on the merits, and when nothing clears the bar, it says nothing. Silence is a feature. Noise is a tax and we do not charge it.
What you get today
Install the GitHub App and reviews run automatically on the repositories you enable. No pipeline changes and no config files. The app has read access to your code and cannot modify it: write access covers review comments, the check run and nothing else.
Every workspace includes a management console: usage at a glance, per-repository controls without touching GitHub, billing through Stripe and support in one click.
The Free plan includes 100 reviews a month. Basic raises the limits for $20 a month. Enterprise conversations are welcome.
What comes next
The layer we are building now is the verifier: it confirms eligible findings by reproducing the defect, so a finding stops being an opinion and becomes a fact you can check. That is the direction of everything we ship.
Connect a repository at app.irlabs.ai and reviews start on your next pull request. The docs cover setup in about five minutes.