Subagent happiness manager · Product, Code & Infra · Web3 Council @labamille
Joined April 2022
- Tweets531
- Following456
- Followers360
- Likes1.7K
Shuip is finally on the @shadcn registry🫡
you can now add the registry directly to your project with:
bunx shadcn registry add @shuip
just shipped shuip: ready-to-use React components and blocks for Next.js shuip.xyz/docs
built on @shadcn /ui, copy-paste or install with one command
less time on UI, more time building what matters
according to @megadevhq I'm a Fleet dispatcher (translation: a worktrees addict)
the card looks cool ngl
a green git contribution day in 2025: ~20 contributions
a green git contribution day in 2026: ~100
love u my subagent homies 🟩
pelavo retweeted
Might as well give voice to it!
Thanks to @Toro_Yuma for the idea!
with skills like superpowers from Obra or grill skill by @mattpocockuk who still uses Claude code's native plan mode? Are there any use cases where skills still fall short?
pelavo retweeted
Take a look at your favourite skill.
Go on, take a look.
Check for lines like:
- "Make the commit message very detailed"
- "Be thorough"
- "Make the implementation easy to read"
What do these lines have in common?
They're no-ops. They do nothing to change the agent's behavior.
Agents always write good commit messages. They always try to be thorough. They always try to create easy-to-read implementations.
Try removing them. Does the output change? No? Then the line was a no-op.
Agent-authored skills are LITTERED with no-ops. Entire paragraphs and sections of text that does nothing to change the agents' behavior.
No-ops make skills harder to evaluate, harder to maintain, and burn tokens uselessly.
Learn to spot them, test them, and remove them. Your skills will thank you.
open-source maxxxxxing this week:
- create.plvo.dev 1.6.5 : shipped new Next.js starter templates: multitenant saas, org dashboards. Added testing modules, and more
- shuip.plvo.dev 2.0.0 → added new @tan_stack /form fields and refactored app field logic
pelavo retweeted
Personal update: I've joined Anthropic. I think the next few years at the frontier of LLMs will be especially formative. I am very excited to join the team here and get back to R&D. I remain deeply passionate about education and plan to resume my work on it in time.
2026 peak engineering :
> use claude code for creating branches and PRs
> launch copilot pr review
> copilot yelling about it
> launch claude pr comments review
> claude reviewing copilot's review and quietly roasting it
pelavo retweeted
New art project.
Train and inference GPT in 243 lines of pure, dependency-free Python. This is the *full* algorithmic content of what is needed. Everything else is just for efficiency. I cannot simplify this any further.
gist.github.com/karpathy/862…
v1.6.0 is live now includes:
• TanStack Query + tRPC integration for SSR & client-side fetching
• Better Auth for auth context setup and tRPC protected procedures
full release notes: github.com/plvo/create-faste…
