Quint

Quint made it onto The Stack Overflow Podcast. A few things @bugarela.com gets into with Ryan Donovan: → why “you’re absolutely wrong” is the answer engineers actually want → the Kubernetes fix Quint rejected before it failed in prod → how writing the spec finds bugs before you run a single tool

More speed, more check marks, more formal verification. None of it matters without understanding. Catch the whole conversation with @bugarela.com on the Type Theory Forall podcast.

We were surprised today by @raymyers.bsky.social replicating @bcherny.bsky.social's formal methods experiment on the Claude Code Agent SDK using Quint Studio (and surfacing a likely bug along the way)! Full video on Craft vs Cruft ↓

New release: Quint v0.33.0 is out! With performance improvements, bug fixes, and support for action properties. You can now write temporal properties about the transitions of your model and check them with TLC ✨ Leaving a star on our repo would mean a lot to us ⭐

We put together a practical reference to formal verification tools: quint.sh/guides/form... If you're getting started, use it to explore the landscape. If you're already an expert, help us make it better.

Don't you love when you're randomly listening to a podcast while working, and you hear yourself mentioned? Great conversation with Tessl on the @ainativedev.io podcast about their dark factory. Quint got a shoutout for helping provide effective guardrails along the way.

The early reviews are in, and we're grateful. Thank you to everyone testing Quint Studio and telling us what you think. And these are only the public notes. The private feedback has been just as generous 💜

The early reviews are in, and we're grateful. Thank you to everyone testing Quint Studio and telling us what you think. And these are only the public notes. The private feedback has been just as generous 💜

Quint Studio gives it to you straight.

Say hello to Quint Studio! Software is full of decisions about how things should work: what happens first, what gets retried, who can do what, and what should never happen.

Agents like to flood. We like to grow.

Remember when programming was fun?

AI changed where the knowledge about your software lives. You might be reviewing code you didn't write and leaning on tests you didn't create.

Coverage and scanners won't tell you. They check the code, not whether the decisions behind your system still hold.

AI changed where the knowledge about your software lives. You might be reviewing code you didn't write and leaning on tests you didn't create.

AI can generate the code and the tests. The pain now lies on knowing which software behaviors those tests actually protect.

The checks you already run catch the obvious problems like a crash or a leak. Nobody has to decide those are bad. Whether your software does what you meant is a judgment call, and only you can make it.

The Quint community has been busy this summer, from bug hunts in production systems to freshly verified consensus protocols. A few highlights we loved seeing:

A special thank you to @mizchi.bsky.social and the wider Japanese community. We are grateful for the attention and excited to see what you all build next.

The Quint community has been busy this summer, from bug hunts in production systems to freshly verified consensus protocols. A few highlights we loved seeing:

gdoteof modeled a job orchestration system and settled open questions about livelocks. github.com/gdoteof/swa...

The Quint community has been busy this summer, from bug hunts in production systems to freshly verified consensus protocols. A few highlights we loved seeing:

Circle shipped Fast Tendermint already formally verified in Quint arxiv.org/abs/2608.13434

The Quint community has been busy this summer, from bug hunts in production systems to freshly verified consensus protocols. A few highlights we loved seeing: