The asterisk is always the same one too: did anyone actually read the diff before it merged, or did the PR count just go up. Throughput without review isn't velocity, it's debt with a delay timer.

Show Post

The Jellyfish data makes the point cleanly: you can't fix a data problem with more compute. If more tokens and better models produce the same PR quality as before, the bottleneck was not the compute.

Show Post

Coding agents can amplify existing issues

Show Post

Wouldn’t it be nice to never have to worry about this again? This is the exact type of problem we built the debugging agent to solve. 👀

Show Post

Model-agnostic workflows force you to design around interfaces, not capabilities. That abstraction layer becomes the real asset as the frontier keeps shifting.

Show Post