davert 🇺🇦 @davert

Prince of Persia on day, developer at night. Author of @CodeceptJS and @Codeception testing frameworks. Currently working on @testomatio to make tests visible! github.com/DavertMik Kyiv. Ukraine Joined November 2008

Tweets

18K
Followers

3K
Following

925
Likes

46K

davert 🇺🇦 @davert

9 hours ago

@RomanHotsiy 10x developer with his agents

0 0 3 46 0

View Details

davert 🇺🇦 @davert

14 hours ago

@RomanHotsiy Yeah pretty big missing feature. I would drop mermaid at all if @excalidraw would have its own markup! I hope you are listening 🙏

0 0 2 21 0

View Details

davert 🇺🇦 @davert

a day ago

@neogoose_btw You write so many sarcastic tweets so only from comments i could find this is not the one Congrats! I would like to use codex more!

0 0 3 114 0

View Details

davert 🇺🇦 @davert

5 days ago

@marcelpociot @typesafeai That's cool people rediscover cheap models My main working horse is gpt oss 20b with 0.07 per m tokens and 1000tps oj groq Really really good for simple agentic tasks

0 0 1 273 0

View Details

davert 🇺🇦 @davert

5 days ago

@VanoDevium @babaikit командувач командування командирів став команидром команди комадування ретельно перевіряйте інформацію!

0 0 2 11 0

View Details

davert 🇺🇦 @davert

5 days ago

@marazmo Звучит прикольно. Это как shared memory layer для гит?

1 0 0 6 0

View Details

davert 🇺🇦 @davert

5 days ago

@adamaltman @typesafeai Yeah, also scratched pr for explorbot but didn't get an endpoint working. Probably i will try it tomorrow...

0 0 0 31 0

View Details

davert 🇺🇦 @davert

a week ago

@letstri Dying on each request

1 0 4 933 0

View Details

davert 🇺🇦 @davert

2 weeks ago

@marazmo @jitel_rf Но ведь ю модем 256к! Пока новость дошла уже позднее время было

0 0 0 17 0

View Details

davert 🇺🇦 @davert

2 weeks ago

Pick cheapest models and run exploratory tests for hours with no interruption ⬇️

testomat.io @testomatio

2 weeks ago

Explorbot uses three model slots: model — reads the page agenticModel — plans and judges visionModel — interprets screenshots Same model in all three, or three different ones. Routing is part of the runtime, so you can balance reasoning quality, latency and cost.

0 0 1 393 0

1 0 2 310 0

View Details

davert 🇺🇦 @davert

2 weeks ago

@LinuxHandbook Still quite popular among testomat.io users

0 0 1 25 0

View Details

davert 🇺🇦 @davert

2 weeks ago

@jeff_behnke_ @OpenRouter @_inception_ai Yeah! Worked good for me as well

0 0 0 29 0

View Details

davert 🇺🇦 @davert

2 weeks ago

@freekmurze @timourxyz You could have asked it to write midi with same effect. Not sure where we need Atra powers here

0 0 1 87 0

View Details

davert 🇺🇦 @davert

3 weeks ago

@christophrumpel While dictating it is hard to go back, switch between sections, expand on points and so on. No way to set file or url references I prefer to type instead

0 0 2 64 0

View Details

davert 🇺🇦 @davert

3 weeks ago

@adamaltman What are sane alternatives? Credit billings? 🤔

1 0 1 20 0

View Details

davert 🇺🇦 @davert

3 weeks ago

@OpenAI For 10$/5 minutes i can do clicking on my own!

0 0 1 30 0

View Details

testomat.io @testomatio

3 weeks ago

"AI testing" can mean a generator, a recorder, a copilot or an autonomous agent. Explorbot is the last one: it plans an exploratory session, drives the staging UI and returns structured tests with video and console evidence. Humans review the run, not every click. Early access:github.com/testomatio/exp…