Lucas Lima

Senior Product Designer. Design Engineer. Brazilian. Agentmaxxing.

[Lucas Lima @_lucasdesigner]

I'm building Moveset, a visual dictionary for design engineering and front-end

each term has an interactive demo and an AI prompt you can give to claude, codex or any other coding agent

https://t.co/4yKGaWhWAv

[Lucas Lima @_lucasdesigner]

where reset

I got places to be, missions to complete, jobs to do, tasks to fulfill, deadlines to crush, targets to hit, goals to achieve, empires to build, destinies to shape, paths to forge, bridges to cross, rivers to tame, contracts to sign, deals to seal, handshakes to exchange, networks to expand, secrets to uncover, mysteries to solve, puzzles to piece, and riddles to crack.

[Lucas Lima @_lucasdesigner]

@raunofreiberg https://t.co/n68GCG0QJB

[Lucas Lima @_lucasdesigner]

the design version of this pyramid is pretty much the same. tokens and components that make the right thing the easy thing at the top, then lint rules and visual regression tests, then rules and skills, and then design review at the bottom, which is the part that doesn't scale.
most design teams i know have it upside down. almost everything lives in review and in a figma file nobody reads

[lauren @poteto]

how do you trust your agents' work more? copy this post/link and give it to your @bot or agent!

1. codebase improvements: this is just good software engineering - better architecture, stronger types and invariants, clear boundaries. agents love to copy what exists and take the shortest path, so make the easy thing the right thing. categorically eliminate problems through code and you won't need to think about it again.
2. static analysis: lint rules, tests, CI, verification (eg TLA+). very token efficient ways to catch issues ahead of time.
3. prose: comments, rules, AGENTS.md, skills - anything that can't be enforced and requires agent judgment. use it minimally, but do use /create-verification-skill.
4. human review: the hardest to scale. at the rate agents land code, it's pretty much impossible to thoroughly review every PR, so spend that effort on the highest leverage point, designing an environment where agents do the right thing by default.

this forms an inverse pyramid: you want most of your effort at the top, least at the bottom. codebase improvements are deterministic, making it more token efficient, and lets you sleep more soundly at night.

install https://t.co/u9xIlN3eN0 and create a new Project:
1. look through my slack/chat messages and code review comments where i've given feedback on PRs, code, architecture
2. use /recall on my past transcripts to learn how i steer agents
3. /poteto-mode use /correct and /architect on past commits to find the most common pitfalls agents fall into, then propose a plan to rearchitect the code i own so it's more agent friendly. answer open questions with prototyping rather than defer to me, and back the plan with real data.
4. before you begin, use /sharpen + /in-your-own-words and confirm this prompt with me

[Lucas Lima @_lucasdesigner]

state machines rock

much better than a pile of frames tbh

[Matt Pocock @mattpocockuk]

I think I'm addicted to getting agents to build state machines

Every time I see an agent has built a state machine (or even just a reducer) I can feel a huge bump in the robustness

And I know that future updates will be trivial and testable

Would recommend for frontend

[Lucas Lima @_lucasdesigner]

Issue tracking UI made with my icons from https://t.co/3mrXSdiQcr

[Lucas Lima @_lucasdesigner]

this is basically what a good research interview does. you think you know what you want until someone asks the third "why" and you realize you don't. I like that agents are making people get grilled on their ideas before building, designers have been trying to sell that for years

[Emil Kowalski @emilkowalski]

Skills like Matt’s /grill-me are great, because by asking lots of questions they help you crystalize your thoughts. They force you to come up with concrete answers to validate your idea. One of the best use cases of AI imo.

You can even go a step further and make a more opinionated skill by feeding it some of your favorite products for example, interviews (think Founders podcast), etc.

[Lucas Lima @_lucasdesigner]

@ridd_design @inflight amazing tool!

[Lucas Lima @_lucasdesigner]

moveset lets you interact with each pattern and compare variations

https://t.co/4yKGaWhoKX

[Lucas Lima @_lucasdesigner]

introducing moveset

https://t.co/C6YLCu74F7

a visual dictionary of design engineering terms. each shows the idea moving, gives it a name, says when to use it and when not to, and comes with a prompt you can give your AI coding agent.

free and open source. new patterns soon.

[Lucas Lima @_lucasdesigner]

I agree with this

AI lowers the execution bar > people start over-designing everything > eventually we circle back and the industry inevitably corrects itself back to core functionality and user experience

the classic pendulum swing of technology and design

[rauno @raunofreiberg]

the wonderful side effect of AI is that once we're collectively fatigued from all the cool "design details" we will finally start to understand design as how it works, not how it looks

[Lucas Lima @_lucasdesigner]

i don't think models are stuck at "no taste" forever, they already pick pretty well when you give them options. what i haven't seen yet is one deciding what the options should be for, and honestly that's most of the work when i'm designing something

[Paul Macgregor @pdotcv]

Agree with this. Humans recognise effort, whether that is in learning a craft, the production process, material sourcing. The less human effort, the less perceived value. Hand made became a luxury signifier in the machine age, in part, because mass production became distasteful.

[Lucas Lima @_lucasdesigner]

We are witnessing software revolution at an incredible pace

What a time to be alive

[Anthropic @AnthropicAI]

New on our Engineering blog: Creating in-house source control.

In May, we merged around 2,500 code changes per day. In September, we merged more than 30,000 per day. As a result, we became bottlenecked by developer infrastructure built for human (rather than agent) scale.

Here, we share how we built a new version control system—and fully migrated our monorepo from git—in five months: https://t.co/wzNkwHs7IN

[Lucas Lima @_lucasdesigner]

merki is made for product UI. sidebars, menus, dashboards, inboxes, productivity, all the stuff that has to stay sharp at 16px

react components, plain SVG, and a Figma library with every icon at 16, 20 and 24 in light and dark

[Lucas Lima @_lucasdesigner]

introducing merki icons

my own opinionated icon library. It's open source and free, so feel free to use it in your projects.

https://t.co/jo20Vp6oBF
https://t.co/gv88jeLuWO
https://t.co/DiUQWvr4qO

v1 is launching with 157 icons. The plan is to launch more packages soon + framework support for Vue, Svelte, plain JS, maybe Solid and Astro. plus the MCP server. After that, React Native, Flutter and Compose.

[Lucas Lima @_lucasdesigner]

@heyimgustavo @vercel parabéns, Gustavo!

[Lucas Lima @_lucasdesigner]

More like they are managing me at this point

[ClaudeDevs @ClaudeDevs]

Claude Managed Agents dynamic workflows are now available in public beta. It's a new type of multiagent orchestration, built for your most ambitious workloads.

A lead agent writes a plan that runs across many agents in phases, combining the results at the end.

[Lucas Lima @_lucasdesigner]

@laurenloprete @mercury hey Lauren is this available for someone based in Brazil? Would love to talk if the location works