Positive users praise the AI agent's long-running code simplification potential and useful prompts, while negative users fear it will delete files or worsen the codebase.
Based on 25 visible X reactions from 62 accounts; directional sample.
Ask a question below.
Published answers will appear here.
@gabriel1 this will end in disaster. it requires everything the models aren't good at: * taste * questioning assumptions * creativity You're gonna wake up to a slop codebase you don't even recognize.
@gabriel1 This is where agents get exciting long-running tasks that improve systems beyond a single interaction.
@gabriel1 “desloppofy” may be the most useful software engineering term invented this year
@gabriel1 thanks for prompt i copy paste now
@gabriel1 And it will be a horrendous result, believe me.
@gabriel1 Top tier prompting right here
i put 5.6 ultra on "desloppofy my codebase and ruthlessly remove any complexity that is not required, and try every feature end to end with @ computer" been working all night and is still working
I think this is how the reflection guy got it to rm -rf $HOME https://twitter.com/gabriel1/status/2075866664864825413
Positive users praise the AI agent's long-running code simplification potential and useful prompts, while negative users fear it will delete files or worsen the codebase.
Based on 25 visible X reactions from 62 accounts; directional sample.
Ask a question below.
Published answers will appear here.
@gabriel1 Top tier prompting right here