Analysis of Opus 5's 'Contrition'

slowmovintarget 17 points 3 comments July 28, 2026
github.com · View on Hacker News

Discussion Highlights (3 comments)

urbandw311er

Does this work? Appreciate honest testimonials from anybody unrelated to the OP or author.

nightshift1

Interesting. I will try some of of those instructions but I don't mind much about the tics. My actual complaint: end-of-turn summaries that are undecipherable unless you watched the session closely. Both Opus and Fable regularly ends hours of work with status updates phrased in crystal clear terms like: "**V2** debug-helper gate (R4d) | Medium-risk | Gate on `?debug=1` URL param at 3 fixture sites — **not** a localStorage flag (`browser_page` clears it every test). 284 debug-API calls ride the fixture, so R4d goes last+alone"

bomewish

Bit frustrating. any kind of benchmark or even a few examples showing that this produces better outputs would have been nice.

Semantic search powered by Rivestack pgvector
15,236 stories · 142,166 chunks indexed