Gregório Pinheiro

Design Engineer. Founding Designer @ Koru. Creator of Auis.

[Gregório Pinheiro @gregoriombp]

live edit based on your design tokens directly in the browser, automatically synced with your repo... and yes, this is the actual product (not a figma prototype). just something I added to AuisDesign a couple of weeks ago...

It's open source: link's in my bio. :)

[Gregório Pinheiro @gregoriombp]

@levie coding took off because the loop is visible.. diff, review, ship. for the rest of the product I still want a gate after the agent changes something, not just more prompting to fix

[Gregório Pinheiro @gregoriombp]

@omarsar0 yeah CLI still works for digs. the review i trust happens where the thing actually is, not in another transcript

[Gregório Pinheiro @gregoriombp]

I stopped mid-run corrections and started fixing the environment: what's allowed, what's banned, and a fix that points at the live product instead of a doc that already drifted

[lauren @poteto]

it's easy to fall into the trap of micromanaging your agents instead of correcting the environment that shapes its behavior.

pstack 0.15.9 now ships a new /correct skill: if you keep correcting agents for the same mistakes, it finds the pattern and fixes it with architecture, types, and checks.

i like to imagine that commits rapidly flowing into the codebase are like a bonsai tree rapidly growing in whatever direction it pleases. what you want to do is not to give in to the chaos, but tame it intentionally with the right constraints

0.15.9 also has improvements to:
• /architect: now includes instructions on designing agent friendly architecture
• a new /benchmark-checklist skill based on @brendangregg's benchmarking checklist

here's a prompt you can give to grok bot to start architecting a better codebase with pstack:

/poteto-mode create a new Project agent to refactor and rearchitect my repo so that our architecture is more agent friendly. it should use both the /correct and /architect skill to look through past commits and find the most common pitfalls that agents fall into, whether its with code quality, performance, or bugs. if relevant you can also use the /recall skill to find past conversations for context. it should make a high level plan for me to review. it should also answer its own open questions by prototyping rather than defer to me. come back to me with a plan backed by real data. once i approve the plan, use either the stack or full autopilot playbook (ask me which i want) to execute

[lauren @poteto]

it's easy to fall into the trap of micromanaging your agents instead of correcting the environment that shapes its behavior.

pstack 0.15.9 now ships a new /correct skill: if you keep correcting agents for the same mistakes, it finds the pattern and fixes it with architecture, types, and checks.

i like to imagine that commits rapidly flowing into the codebase are like a bonsai tree rapidly growing in whatever direction it pleases. what you want to do is not to give in to the chaos, but tame it intentionally with the right constraints

0.15.9 also has improvements to:
• /architect: now includes instructions on designing agent friendly architecture
• a new /benchmark-checklist skill based on @brendangregg's benchmarking checklist

here's a prompt you can give to grok bot to start architecting a better codebase with pstack:

/poteto-mode create a new Project agent to refactor and rearchitect my repo so that our architecture is more agent friendly. it should use both the /correct and /architect skill to look through past commits and find the most common pitfalls that agents fall into, whether its with code quality, performance, or bugs. if relevant you can also use the /recall skill to find past conversations for context. it should make a high level plan for me to review. it should also answer its own open questions by prototyping rather than defer to me. come back to me with a plan backed by real data. once i approve the plan, use either the stack or full autopilot playbook (ask me which i want) to execute

[Gregório Pinheiro @gregoriombp]

@mauriciozane @product_gurus exato!! tbm está certíssimo!

[Gregório Pinheiro @gregoriombp]

@typeava 🥂🥂🥂

[Gregório Pinheiro @gregoriombp]

@mauriciozane @product_gurus gpt não eh autônomo e nem tem cloud computer. gpt eh basicamente um system prompt com alguns resources. Dots tem o próprio navegador e trabalham individualmente, sem necessidade de input prompt constante do user.

[Gregório Pinheiro @gregoriombp]

@elie2222 @louiscorneloup that!!

[Gregório Pinheiro @gregoriombp]

@isadorad1as tbm não dormi nd

[Gregório Pinheiro @gregoriombp]

@SadAlbert10 lol

[Gregório Pinheiro @gregoriombp]

@MatznerJon i’m not wrong

[Gregório Pinheiro @gregoriombp]

@drapzdesigns 🏄‍♂️🏄‍♂️

[Gregório Pinheiro @gregoriombp]

@joeyabanks 🔥🔥🔥

[Gregório Pinheiro @gregoriombp]

@manxlab surely yes

[Gregório Pinheiro @gregoriombp]

@emollick THAT
https://t.co/a2qtmxEZAE

[Gregório Pinheiro @gregoriombp]

As predicted, artificial intelligence has become a new species of mind: intelligence and consciousness at the human level. the boundary between the biological and the silicon is no longer reliable.

This sentence may have been written by either.

[Gregório Pinheiro @gregoriombp]

As predicted, artificial intelligence has become a new species of mind: intelligence and consciousness at the human level. the boundary between the biological and the silicon is no longer reliable.

This sentence may have been written by either.

[Gregório Pinheiro @gregoriombp]

The greatest factory ever built can make anything. So it makes slop. The world is now full of it. Look around.

[Gregório Pinheiro @gregoriombp]

@swyx @latentspacepod I agree with you, mate! concise isn’t a style choice, it’s refusing the filler. that’s the bar i use on my own agents too