I think the improvement on how it codes is pretty much represented correctly by the benchmarks (a nice bump, but not some crazy leap)

But where it really shines is in how NOT lazy it is. Fable requires less hand-holding. And I can understand how someone who uses Claude-Code sparingly and with very focused prompts would not see a lot of improvement there.

But simple example: if you ask Opus to do a review of the codebase (with a short prompt and not too much guidance), I've had it basically read the `git log` output, do a simple `ls` and have it declare "Everything looks great! No problems found!", when Fable really does what you would expect it to do.

And you might think: "oh, so it's just capable of handling crap prompts?", well sure. But even if you make THE PERFECT Opus plan (a plan that would take many turns/hours to finish), Opus will fake out, say everything is done, and then you see that half of the plan was deferred, half of the functions are ridiculous stubs, ...

If you give the same plan to Fable, it'll just DO IT. And it WILL get it done. And in the end it'll tell you "Oh, I also found 30 other bugs and I fixed all of them properly" (where Opus would have started crying, or WORSE, worked around the bugs)

▲

esperent 2 hours ago | parent | next [-]

> Opus will fake out, say everything is done, and then you see that half of the plan was deferred, half of the functions are ridiculous stubs, ...

Doesn't Claude Code have a /loop command? Give it a message to keep it on track overnight, send every 20m, make it track progress in a doc, reread the doc after every loop. I've found this works well for a certain class of problems, most importantly where the actual work is getting done by very narrowly focused batches of subagents, with the main session just coordinating and keeping the doc updated.

	▲	hadlock an hour ago \| parent \| next [-]
		They added a "/goal" command which I guess spawns a supervisor agent process that checks to see if your goal statement has been achieved (e.g. "/goal complete tasks 1-250 of plan.md") I've been pretty happy with it but I rarely use that workflow. Most of the time I give it a 3-6 step prompt and come back in 20 min and the first two were done and I get a summary "up next is to complete the next steps" which.... Opus 4.6 didn't have this problem. 4.8 feels like a cost cutting measure, or maybe it's just tuned poorly for my specific workflow (multi-repo system integration)
	▲	whatinthe67 2 hours ago \| parent \| prev [-]
		For optimizations or proofs I suppose? Wouldn't know why else you would do something like that.

▲

flatline 3 hours ago | parent | prev [-]

I think the parent comment stands - I’ve asked Opus to do a review of DeepSeek’s test suite and told it a couple things I wanted it to look for, and it did a very thorough review of the tests and picked out a reasonable number of gaps and tautological tests. It’s a mix of prompting/instructions, the agent harness, and random chance. The model is not wholly irrelevant but IMO increasingly so.