Aaron Helton 5 days ago
The argument against this practice seems to be that since the coding agent can’t really be held responsible for its behavior, nobody should, not even the developer who uses it on code without checking what the output will be.
add a skeleton here at some point