Загрузка видео...
Не удалось загрузить видео
this prompt turns your grok bot/codex dot into a 300 IQ personal assistant.... it studies your chats,worklfows, and memories then finds work you can fully automate paste this into your agent👇 "Review all my Claude, Codex, and Grok Bot conversations you can access, along with my local workflows and... show more
28,508 просмотров • 6 дней назад •via X (Twitter)
Комментарии: 38

I ll try this

yes let me know how it works out

codex dot flagged my buried slack workflows nobody used anymore

this is very useful. it allowed me to rearrange my threads as well

the real value is not the model itself, but how well it understands your workflow. automating decisions is where things get interesting.

I saw you made one for avid 🤣

i saw it. it's insane

wow i need to give this prompt a try inside my grok bot

You need to try it within any AI tool that you have

@Av1dlive make sure it doesn’t automate stuff that needs a human touch. Learned that the hard way.

Yes so I have something called a taste.MD which allows me to define stuff which is out of AI reach

@Av1dlive that's interesting! How does taste.MD work for you?

amazing prompt bro, i love the way how prompt is structured

my agent genuinely started understanding me better and handling tasks but mine's built in claude, so it works for him too

Yes you can use this prompt for that as well

First it was grok bot, now it’s dots. Agents are making things easier to manage

new find!! ranking by setup efforts is the smartest line here

basically asking your AI to find its own work

That sounds like a game-changer! I love the idea of having a super-smart assistant that actually understands my work.

These prompts are super valuable. That's exactly what I was looking for! I'm already hooked on Grok, and I want it to keep getting to know me better so it can run like my 24/7 personal executive assistant.

Nice post

thanks bro

先把聊天记录喂给它,隐私这关就够我纠结了

Turning chat history into an automation backlog is a pretty clever workflow

hmm makes sense to let the agent go find its own work

Yeah, it often seems that with these models we are the bottleneck

i think the best audit is the stuff you keep redoing by hand every week, and the agent can only see that if it has logs of what you did, not just chats. otherwise it only knows what you talked about

the context window management here is going to fail silently on long threads

"review all my conversations" is a lot of access to hand over in one paste. i'd try it on one week of chats first and see what it digs up

lemme try

This is like a workflow audit with a time-savings column: the useful part isn’t the assistant calling itself smart, it’s ranking automations by payoff and setup cost.

so huge share homie

thanks bro

This is the best solution for working with Grok Bot

ranking by time saved and setup effort is a great idea it shows where to start and not just a list

Need to put my hands on

this seems risky, relying solely on past chats might overlook new insights

step two: train the repeatable parts instead of re-prompting forever


