Messing with AI. Opinions earned, not assumed.
Barcelona
- Tweets 4,940
- Following 1,783
- Followers 567
- Likes 858
Agree. I've been having the same feeling for awhile
As it stands now, here is my hottest take:
A software factory is actually an applied automation practice in the actual product. It's not some external thing. It's a product pattern where the product builds the product inception style. Any external system or dependency should be internalized into the product to enable the product to build the product.
If you can't build the product in the product, from the product, then you are missing the mark.
I don't know how to explain this more clearly at this stage. It's a bit mind-bending, but if you understand meta programming, macros and understand the Factorio reference that a factory should build the factory so it can build the factory. It might be a little bit easier to follow along....
>Your product is the factory<
ghuntley.com/rad
1
72
jev is the new thing
everyone is building jevs
so obviously I wanted my own
technically, a jev skips the whole token-by-token generation thing
it reads the logits straight from prefill and picks from there
which is kinda cool because you can cache the prompt once and run a bunch of questions in parallel
2
2
75
...anyway, I built one
you can try it here ↓ github.com/mpuig/system-one
1
1
20
coding agents are pretty good at writing Mermaid. making the diagrams readable? sometimes they need a little help 😅
just shipped Beautiflow 0.9: a CLI + agent skill for Claude Code, Codex, and Pi.
beautiflow.cc
4
1
101
also gave architecture diagrams some love: clearer labels, better connectors, and fewer awkward overlaps.
not just AWS—GCP, Azure, or whatever you’re building.
open source, no extra model API keys. give it a try 👇
github.com/mpuig/beautiflow/…
24
Marc Puig retweeted
[1] software factories are super real but we need to be realistic. the factory aspects haven’t been cracked yet, the [3] innovators are toying around with discovering practices and pieces.
if someone is selling you a factory rn and they aren’t in the super small cohort (ie. they are likely startups created in last six months!) or circa <10 people in the who’s-who, who have been trying (and realistically failing) over the last two years.
they are selling bullshit.
it isn’t cracked but it’s a puzzle that’s being solved daily. the best practices right now are to build abstractions/pieces needed which entails solving [2] non-agent topics.
[1] almost lights out.
[2] sandboxing, monorepo, reproducible builds, ci/cd, identity/secret management, smashing corporate friction in the realm of devex for agents.
[3] at home, in your homelab, you should be cracking on this problem space in your free time if you want a super fast promotion and red carpet service at your next interview. highest roi you can have right now is learning the entire damn stack, automating it and showing it at your next interview/doing a recorded talk at a meetup.
61
65
872
138,477
Been buried in work travel for weeks, but I finally got a window to update copycopy.app
It’s now running Gemma 4 E2B QAT. Small update, big improvement.
4
124
Marc Puig retweeted
New blog: Building agents that reach production systems with MCP.
When should agents use direct APIs vs CLIs vs MCP? Plus patterns for building MCP servers, context-efficient clients and pairing MCP with skills.
claude.com/blog/building-age…
92
315
3,279
481,885
CopyCopy: on-device AI for your clipboard.
Summarize, rewrite, translate.
Nothing leaves your Mac.
CopyCopy.app
59
Gemma 4 is the first release in a while that makes the local-first future feel closer.
I’ve been pretty convinced for a while now that local-first is where this is going.
What feels different now is that it’s starting to make sense all at once: technically, economically, and culturally.
1
50
I ❤️ local inference getting faster and cheaper on real hardware. Once developers can benchmark what actually runs on their own machines, cloud-only assumptions start to break down fast.
Ollama is now updated to run the fastest on Apple silicon, powered by MLX, Apple's machine learning framework.
This change unlocks much faster performance to accelerate demanding work on macOS:
- Personal assistants like OpenClaw
- Coding agents like Claude Code, OpenCode, or Codex
83
Using OpenAI Symphony to push further on harness engineering.
I wanted agents to pull work directly from GitHub Projects, so I built the GitHub tracker adapter:
👉 github.com/mpuig/symphony/tr…
1
1
118
i built an agent skill for powerpoint decks. both the slides and the tooling were not straight forward.
the interesting part about writing a skill for ppt slides is what i learned making tools work when the user is a language model.
llms don't read "--help". they hallucinate field names. they'll dump your entire schema into context and wonder why they can't reason anymore.
what actually works: runtime introspection over docs, structured errors over helpful messages, validation before mutation, progressive disclosure over
dump-everything.
this isn't a slides thing. it's a pattern. every tool you build for agents will need this.
1
3
70
i think it's about to be everywhere so I created a site about Agent Development eXperience
agent-dx.com
1
41
I built an open-source agent skill that generates PPT decks from a single prompt.
Yes, Claude Cowork can do this natively now. I know. I did it anyway - because I wanted it to work in any agent, not just one.
Claude Code, Pi, Codex. Your pick.
👉 npx skills add github.com/mpuig/agent-slide…
2
3
147