Engineer Guided AI-Driven Software Engineering | Agent-Based Systems | Spec-Driven Development
Saint Petersburg FL
Joined July 2009
- Tweets9.5K
- Following2.3K
- Followers640
- Likes3.2K
Pinned Tweet
We need more standardization on managing ai coding agent skills, plugins, and agents. Especially with configuring agent llm providers, mcps, lsps, skills etc. It shouldn't be a chore to switch harnesses and manage all this. There's too much vendor lock in
Using @apple invites for my wedding/rehearsal - @AppleSupport it would be great if i could post multiple images (flyers) that are included with the event besides a background image, should be more than 5 links - also it should integrate with spotify, tempted to make a competitor
Michael L. Benin retweeted
Absolutely incredible showcase of what can be built with @neondatabase Realtime! 🚀🎉
It's not just cursors that sync, the full state of the document is a set of live queries straight from Postgres - that's all it take to keep all users in sync!
Introducing orcanva.app
A fully open source and 100% free multiplayer Canva alternative.
I make all my thumbnails here. It's simple and easy to use.
You can jump into public designs with other people, no signup needed. Perfect for making memes live.
Michael L. Benin retweeted
We’re doubling down on open source.
Today we're making PowerSync's entire core open source by moving our remaining source-available (FSL) repos to Apache-2.0
Our hypothesis: less legal friction → more adoption → a stronger business → more investment in open source.
Read 👇
What if coding agents didn’t just remember: they learned? @github Copilot CLI (harness) captures traces, @Vectorizeio Hindsight preserves memory, and @modal trains/evals adapters from verified merged work. Harness evolves → memory compounds → model improves. Own the workflow.
Michael L. Benin retweeted
The Local-first manifesto has 7 ideals. I still believe in 6 of them...
However, I don't think "the network is optional".
The "ideal" is that we have robust and stable connections everywhere.
I always used to use the example of traveling through a tunnel on a train, loosing your connection, as to why Local First was important.
Now, almost all my work is happening on remote long running sandboxes or machines - even personal stuff.
Some may say that local AI is the solution.
But what if I'm on my phone and I had an agent running on a multi-day development session?
Or a research agent investigating something online?
Or a scheduled agents checking in and looking to see if the gig tickets I wanted are available?
I can work on none of that without the network.
The network is not optional.
Focused on using harnesses to get work done.
Not making my own harnesses, making my workflow portable between harnesses.
I'm focusing on avoiding vendor lock in at all levels.
I've seen it too many times where a tech you invested time in suddenly changes from under your feet
Michael L. Benin retweeted
The most difficult part of Software Engineering was never writing the code.
It was decision-making.
Michael L. Benin retweeted
No, WebMCP isn't replacing MCP.
MCP connects agents to your services.
WebMCP connects agents to your website while it's open and the user's in the loop.
Here's how they fit together ⤵️
Michael L. Benin retweeted
10 repos de GitHub para agentes IA que más han crecido en Octubre:
1. Hindsight
Memoria para tus agentes que aprende de lo que hacen.
+28.5K estrellas en el último mes.
github.com/vectorize-io/hind…
2. Paperclip
Una app para gestionar a todos tus agentes en el trabajo, como si fueran un equipo.
+24.8K, ya pasa de 105.9K.
github.com/paperclipai/paper…
3. ECC
Skills, memoria y seguridad para Claude Code, Codex, OpenCode y Cursor.
+22.5K, va por 286K.
github.com/affaan-m/ECC
4. Impeccable
Un lenguaje de diseño para que lo que diseña tu agente no parezca hecho por una IA.
+15.1K.
github.com/pbakaus/impeccabl…
5. FreeLLMAPI
34 proveedores gratis y 635 modelos detrás de un solo endpoint. Si uno cae, salta al siguiente.
+11.2K.
github.com/tashfeenahmed/fre…
6. OpenShell de NVIDIA
Un entorno seguro y privado donde dejar trabajar a tus agentes.
+8.3K.
github.com/NVIDIA/OpenShell
7. Plugins oficiales de Claude Code
El directorio de plugins que mantiene la propia Anthropic.
+7.2K.
github.com/anthropics/claude…
8. CLI-Anything
Convierte cualquier programa en algo que tu agente puede manejar.
+6.4K.
github.com/HKUDS/CLI-Anythin…
9. text-to-cad
Tu agente diseña piezas en 3D a partir de texto.
+6.1K.
github.com/earthtojake/text-…
10. Knowledge Work Plugins de Anthropic
Plugins para usar Claude en ventas, legal, finanzas y marketing.
+5.8K.
github.com/anthropics/knowle…
Todos open source y gratis.
Michael L. Benin retweeted
We can fix smart TVs too?!
OmarchyTV is out in beta
An #Omarchy style launcher for Android TV: themes, Waybar style bar, live stats and a terminal-style screensaver.
It comes with the omarchy-tv plugin, so your Omarchy PC can push themes and share its screen to the TV.
Based on tv-vibehack by @gathint. Built for Omarchy by @dhh ❤️
It's on GitLab for now (my GitHub account is still stuck in review):
gitlab.com/piermark/omarchyt…
Wallpaper: Liftoff by @tcballard
Michael L. Benin retweeted
MiniMax M3.1 Flash goes toe-to-toe with Opus and GPT on coding at ~25x lower cost.
This was the model behind Space Bunny, the stealth model that’s been topping usage leaderboards recently.
Free in Cline now!
MiniMax M3.1 Flash: Space Performance + Bunny Speed
🚀🐰 Yes, Space Bunny - which has topped global leaderboards with 200T+ tokens - was an earlier version of M3.1 Flash.
💰Redefining the performance economics of agentic AI. M3.1 Flash achieves frontier-level intelligence across key tasks such as Cowork (Automation-Bench: 62%; ALE: 30%), Coding (DeepSWE: 74%) and Visual Reasoning (BabyVision: 91%), at 1/3 the cost of M3 and 3x the speed.
API pricing (flat rate): $0.15 / $0.60 per 1M input/output tokens, with cache reads at $0.015 / 1M tokens.
Model weights will arrive on Oct 23.
API: platform.minimax.io
M Plan: minimax.io/m-plan
MiniMax Code for desktop: agent.minimax.io/download
Michael L. Benin retweeted
in the next version of Porffor, Test262 compliance leaps to 96%! this marks a milestone where Node compat and performance finally become major blockers
ALT "porffor passes 96.05% of test262" with a graph showing progress over time
Michael L. Benin retweeted
There is a bunch of "TypeScript compiled to native" projects about, but this is the OG and the one to watch!
Oh, and it's not just TypeScript, it's all of JavaScript!
It's wild how google open sources something that's already been done by a ton of devs yet because google authored it, it gets more attention
Michael L. Benin retweeted
Cloudflare open-sourced the security audit skill that seeded its own vulnerability hunting harness. Your coding agent can now run it on your code.
How it works, in 6 phases:
→ maps your codebase
→ hunts for bugs area by area so nothing gets skipped
→ sends every finding to a fresh agent that didn't find it, to confirm it's real
→ writes structured results
→ verifies each record again, independently
→ writes the final report
You get a findings file split into confirmed, needs validation and rejected, so you fix the real problems first.
Install:
npx skills add github.com/cloudflare/securi… --skill security-audit
Then open your agent in your project and type: "security audit this codebase"
MIT licensed, 31.3K stars.
