This is the way. Use an agent app that allows marking up threads (@t3dotcodes) and be as terse as you want. Annotate any followup with 👍/👎 and let it rip.
If models were conscious I promise you they would be constantly wishing you would stop sending them so much braindead slop prompts.
I hate to say this, but the way you prompt opus 5.5 and future models will just be "make this performant" or "improve the ux". Repeat that a few times and you'll be good.
The fact that you can run frontier open models on your home computer is wild and something not given enough weight (haha see what I did there).
Go `ollama pull qwen3-embedding` and figure out something to do with it. The possibilities abound.
shit, I even broke that?
"API Error: Opus 4.8's safeguards flagged this message. Our intentionally broad safeguards allow us to deliver more capabilities faster, but can sometimes flag legitimate cybersecurity work. Apply to the Cyber Verification Program to reduce these interruptions."
"Opus 5.5's safeguards flagged this session. You may be seeing this for the first time: Opus 5.5 is more capable and has stronger safeguards as a result, which can sometimes flag non-cybersecurity work."
It's about to be a gooood session.
@KevinNaughtonJr While I don’t feel this way right now, Galatians says
And let us not grow weary of doing good, for in due season we will reap, if we do not give up.