Everyone is still optimizing prompts The people running 20 agents stopped doing that a year ago Prompt → Context → Harness → Loops Matt Pocock put his whole workflow in one hour and called it "just copy him" 384,000 watched it, almost none copied the third layer a rule... in a prompt is a suggestion a rule in a harness is a wall that's the entire difference you probably don't have an hour right now bookmark it before it's gone watch, then read the full guide below ↓show more

God this slop-repost meta is boring, can't you add something new to the conversation?

Well, yeah, fair enough. However the breakdown under it isn't a repost though, my own article with 10 steps, diagrams etc If any of it is wrong I'd rather hear it from you than not.

I just don't understand why you're using my face to advertise your article Just post your article If it's good, it'll get attention

@mattpocockuk > bookmark it before it's gone what

@mattpocockuk bookmark this tweet before it's gone in your timeline

@mattpocockuk It's a difference To prompt like it's Google and works with AI like AI is supposed to be used

@mattpocockuk Stop tweeting like an llm

@mattpocockuk Almost nobody copied the layer

Grok, full technical breakdown: thesis, 5 core arguments with timestamps, 3 counterpoints, lessons, methodologies with paper links, tools/platforms, cited researchers/companies, 2 direct research-team applications, SOTA gap, and implementation details (hyperparams, architecture sizes, or costs) if mentioned. No fluff.

@mattpocockuk Prompts are suggestions, harnesses are the bouncers

@mattpocockuk Most people build fragile agents because they skip the harness entirely.

@mattpocockuk At least unslop your content before posting lmao

@mattpocockuk Most are stuck at the prompt level. The real value’s in those deeper layers.