I've been watching Mo build this for what seems like an age but super excited to have been onboarded and trying out the beta. I absolutely love being able to bring my agents in to work with me, and the threading is perfect. Can't wait to see Inline Chat evolve 🤌
the thread-based approach makes sense for keeping agent output out of the way of the human conversation, but I'm curious what happens when a human jumps into a thread mid-way through an agent's task and changes direction. does the agent see that as new context in the same thread and course-correct, or does someone need to kill it and re-prompt from scratch? that moment feels like where a lot of "agent working alongside teammates" tools end up feeling clunky in practice.
Forward
I've been watching Mo build this for what seems like an age but super excited to have been onboarded and trying out the beta. I absolutely love being able to bring my agents in to work with me, and the threading is perfect. Can't wait to see Inline Chat evolve 🤌
There
@niccoates Thank you Nic for all your support and following along over the years! ❤️
Dial
the thread-based approach makes sense for keeping agent output out of the way of the human conversation, but I'm curious what happens when a human jumps into a thread mid-way through an agent's task and changes direction. does the agent see that as new context in the same thread and course-correct, or does someone need to kill it and re-prompt from scratch? that moment feels like where a lot of "agent working alongside teammates" tools end up feeling clunky in practice.