Загрузка видео...

Не удалось загрузить видео

На главную

I pointed the new Claude Managed Agents dynamic workflows at 18k lines of code. > 21 proven bugs. > 5 minutes. > $8.36, on my Claude Max API credits. A lead agent writes a plan, runs it across many agents in phases, then merges the results. Try it in...

11,788 просмотров • 1 месяц назад •via X (Twitter)

Комментарии: 23

Фото профиля Jason Haugh
Jason Haugh1 месяц назад

This is my Saturday

Фото профиля Dan McAteer
Dan McAteer1 месяц назад

Enjoy it!

Фото профиля Rick Ye
Rick Ye1 месяц назад

faster usage burn is the key to success 🔥

Фото профиля Dan McAteer
Dan McAteer1 месяц назад

Guess it depends on your definition of success 😂

Фото профиля EvalGuys
EvalGuys1 месяц назад

now the fun part: finding out how many of the 21 the author calls "by design"

Фото профиля Danny Veiga
Danny Veiga1 месяц назад

Cheapest bug bounty program ever run lol

Фото профиля Dan McAteer
Dan McAteer1 месяц назад

Worth every Penny!

Фото профиля Anton
Anton1 месяц назад

What happened with the sol-advisor, going to update it so it runs sol 6.1?

Фото профиля Hank Yeomans
Hank Yeomans1 месяц назад

that /claude-api is to die your api access?. if So my mythos 5.1 access is via api so this would be useful

Фото профиля Dan McAteer
Dan McAteer1 месяц назад

I think it basically gives Claude the context to build with the API

Фото профиля Blanc
Blanc1 месяц назад

what's the false positive rate? an agent emitting 21 findings in 5 min is cheap; 21 findings that each reproduce with a failing test is a result.

Фото профиля EgeTheCeo
EgeTheCeo1 месяц назад

21 bugs in 18k lines for like $8 is wild ngl how many held up once you actually read through them? and did the merge step ever choke when two agents touched the same file

Фото профиля Adrian Lin
Adrian Lin1 месяц назад

8 bucks to find 21 bugs in five minutes. my code review process is just me squinting at the screen until i feel vaguely guilty.

Фото профиля 安叫兽|Bird🕊️ 🔶 BNB
安叫兽|Bird🕊️ 🔶 BNB1 месяц назад

有点好奇这批 bug 里有几个是现有测试没兜住的

Фото профиля Michael Cohen
Michael Cohen1 месяц назад

awesome.

Фото профиля Abhishek Ray
Abhishek Ray1 месяц назад

thats not too bad - how bad were the bugs? also did it run opus for the agents?

Фото профиля Daniel Kim
Daniel Kim1 месяц назад

21개 버그가 실제 검수에서 얼마나 남는지와 에이전트 단계별 비용을 같이 기록하면, 병렬 실행의 효율을 더 정확히 비교할 수 있겠네요.

Фото профиля Dui Toledo
Dui Toledo1 месяц назад

love this for work that's parallelizable and can run in isolation. but dynamic workflows don't solve what i'm still figuring out with 50-100 agents: how to keep them building features and real value without stepping on each other's toes, especially when different people are running them

Фото профиля saksham dogra
saksham dogra1 месяц назад

the interesting bit isn't the agent count, it's the orchestration. 21 real bugs in 18k lines from one automated pass is honestly wild — results like that make the "agent swarms are just token waste" take a lot harder to hold

Фото профиля Chloe Bennett
Chloe Bennett1 месяц назад

a lead agent that plans then fans out and merges is the workflow that finally feels like it scales past one chat window

Фото профиля catman
catman1 месяц назад

It’s like sending a lead engineer to scout the codebase while specialists inspect separate areas, then reconciling the findings into one review.

Фото профиля CSI
CSI1 месяц назад

Use your own agents outside of Claude Code. They can collaborate and share permission-based access to files behind your own Blackboard.

Фото профиля Cotal
Cotal1 месяц назад

Does each finding come with a runnable repro or regression test? Would be useful to see one of the bugs from discovery through to a verified fix.

Похожие видео