Already made a harness for Claude to make R/W plans, not write once like they are usually implemented. They can modify themselves as they work through the task at hand. Also relying on a collection of patterns for writing coding task plans which evolves by reflection. Everything is designed so I could run Claude in yolo-mode in a sandbox for long stretches of time.
I think that a curl wrapper could serve as a universal integration with arbitrary services (as long as they have a public API), saving you a lot of setup complexity and context space. Authentication is the hard part; ideally, you shouldn't even need an OAuth intermediary. Which I think is doable: https://news.ycombinator.com/item?id=46876636. It's still a bit of an experiment, though.
Whatever we do now to "steer" the model to do the job, my 5 cents, it will all get sucked into the model itself; skate where the puck is going as they say, and relentlessly focus on user experience and the overall product, that's how you get something like Granola.
I was happy enough to use Vimeo for several years, then made an account to be able to watch videos all of a sudden, and just recently when I logged in they've asked me to also verify (?) my identity with an ID. Yeah, not doing that.
I can't parse this story. "rei" stopped working and you asked "rei" or "clawdbot" to help your troubleshoot? Are you using both? Whos is 'we' in the "we fixed it ourselves" substory?
Clawdbot is the software, they installed their own instance of it and named it "rei". So an instance of Clawdbot named rei helped them to fix a problem in Clawdbot/rei they observed.
reply