dev

working on all sorts of things 🦄🦄

[dev @sa_han47]

@Baconbrix bacon-bench for sure!

[dev @sa_han47]

@agupta This part of the video really did something to me @theo !!

https://t.co/7WcwtxaL8g

[dev @sa_han47]

hey @threepointone,
How would you map the Claude managed agents primitives to the agents framework primitives?

ex: is there a vault(oAuth token storage and refresh) concept in CF?

[dev @sa_han47]

@zeeg Oh no!!!

[dev @sa_han47]

@paulg Are you travelling at the moment in New Zealand? I swear I saw you

[dev @sa_han47]

@threepointone hey! Similar to how Anthropic managed agents has "vault" - does CF Agents have a similar primitive to store secrets in?

[dev @sa_han47]

@JayHoovy bro, please do more longform company videos - miss watching those on youtube

[dev @sa_han47]

@PriyankaPhatak please help answer this https://t.co/f1gMxD0jFs

[dev @sa_han47]

@bcherny @PriyankaPhatak Do you know if the "Managed agents" team will add the ability to setup cache-diagnosis? I do not see a way to set this up

[dev @sa_han47]

@lgrammel Hey Lars!
Is there any guidance on how "Claude managed agents" could be hooked up to useChat() sort of experience? Is that even a thing? My company is keen to build an agent that could be called from anywhere - just wondering if you've seen this done before?

[dev @sa_han47]

@fayazara question: how do you build this stack on top of https://t.co/xS4YeiGrXc? Having a hardtime understanding when to use a tool like useChat hooks vs an agent that run async, can the same agent be used but the "UI" is different?

[dev @sa_han47]

@ClaudeDevs https://t.co/f1gMxD0jFs

anyway to solve this?

[dev @sa_han47]

@bcherny @PriyankaPhatak Do you know if the "Managed agents" team will add the ability to setup cache-diagnosis? I do not see a way to set this up

[dev @sa_han47]

@JayHoovy Can we please have a new episode of the "Stan" update!! I love watching those videos of you guys building this agent with the team

[dev @sa_han47]

@bcherny how do you setup a orchestrator to manage the work and verification of the results?

[dev @sa_han47]

@bcherny @PriyankaPhatak @TheAmolAvasare @lydiahallie any help here please?

[dev @sa_han47]

@bcherny @PriyankaPhatak Do you know if the "Managed agents" team will add the ability to setup cache-diagnosis? I do not see a way to set this up

[dev @sa_han47]

@ambricken whose tokens does Claude use when tagged?

[dev @sa_han47]

@RhysSullivan if someone needs AI, wouldn't they just connect this to their ChatGPT + Claude?

[dev @sa_han47]

@BrendanFalk @usehercules @Steve8708 check this out! Similar to what you did

[dev @sa_han47]

@paulg It has to be some form of "non-invasive"(I should be able to disconnect and go on a nature walk) neuralink

[dev @sa_han47]

@ctatedev how does this compare to something like https://t.co/XjutGaSDCE?