The current aifr --format=text log output is not pretty and it's hard to tell apart commits.
How well optimized are models for reading regular git log output? How about with --oneline?
How often do coding agents use git log --pretty=tformat:... and friends?
Some token parsers prefer an XML format. Is git log a mode where XML output makes sense, token-wise?
The current
aifr --format=text logoutput is not pretty and it's hard to tell apart commits.How well optimized are models for reading regular
git logoutput? How about with--oneline?How often do coding agents use
git log --pretty=tformat:...and friends?Some token parsers prefer an XML format. Is
git loga mode where XML output makes sense, token-wise?