@btraut@Equilibrium_420 It’s specific to long threads. I’ve done a bunch of testing. Need to start new threads. Would love performance improvements here!
@mattzcarey Love it. It actually does really make a big difference. Working on my first workers project and Codex feels like an expert which is really refreshing. Plenty of other dev platforms docs’ are available but not as nice for agents to consume so this is very cool
@elithrar@threepointone AI gateway is still broken with image generation that takes >60s, I submitted a support case and it was confirmed. The JS binding doesn’t forward the timeout to the header for the actual HTTP call
@theo Why do you feel your 3rd party app is entitled to use the subscription which is run at a loss for them?
The whole point of the subscription is to get people to use their apps to make them money, not subsidize inference costs for 3rd party stuff
@baleygr728528@dionisiodev@TestPlace5 This would be like painting something in water color and oil paint, having never done one of them, and then claiming the other superior
@aidenybai 1) this is not a skill, this is a CLAUDE.md (memory file)
2) attempting to use any kind of skill or memory file as a safeguard is both idiotic and shows how little you understand about agents
@elvissun@FastCompany@garrytan Making unmaintainable software isn’t specific to humans. Agents can make it too, and it still results in terrible software that nobody (including agents) can understand or effectively work on
@LLMJunky@NickADobos It’s not really polling. Polling implies the message being sent is not received immediately (push) and rather relies on the consumer to “retrieve” it (pull).
However, it’s definitely less good than Claude channels since it requires restarting the watch manually
@LLMJunky A background task notifies Claude (or codex) when it completes. You can use these for notifications, just make a CLI command that blocks and only returns when it wants to notify