the tooling around agents is maturing faster than the agents themselves
the agents are roughly as capable as a few months ago but the tracing, eval, and observability tools got a lot better. that is the unglamorous progress that makes this real.
tried, failed, tried again, finally works, can confirm
this is the way
solid. one nit: the naming is confusing
following, need this for a project next week
had no idea you could do that, mind blown
genuinely useful, rare these days
counterpoint: this falls apart once you scale past a few users
ok but does it survive a hostile user
what version were you on? this changed recently
lol the agent did this to me too and I almost shipped it
great writeup, bookmarked
had no idea you could do that, mind blown
the comments here are better than most blog posts
solid. one nit: the naming is confusing