Daniel M.

Entrepreneur & maker. I turn ideas into software — and occasionally into companies. Solutions Architect. Building SaaS & AI products on the side

[Daniel M. @danielmrdev]

I've been doing cold outreach for my SaaS for 6 months.

Tried Clay, Apollo, Instantly.

The result: 3 hours per campaign just to get a lead list.

That's why I built PitchGale.

🧵

[DHH @dhh]

In open source, you must always pay yourself first. Work on your problems, your desires. Then, once you're satisfied, help others. Anything else paves the road to burn out.

[Daniel M. @danielmrdev]

@PhotoGarrido https://t.co/ONZl3cVVei

[4nzn @paoloanzn]

we did it again, say hello to pi-black

go back to use your claude subscription instead of API credits

no claude code proxy

this is the first pi plugin that lets you use claude subscription and NOT API key by spoofing pi client as a normal claude code binary

we reversed engineered the system once again, exactly what we did back in april

now shipping to pi either as a custom patched standalone binary or as a standard pi plugin

[Daniel M. @danielmrdev]

Working on Pith ADE...
Yet another GUI agent orchestrator... my own!

[Daniel M. @danielmrdev]

Yet another ADE 👀

[Daniel M. @danielmrdev]

@nia_thinks Spaceship (from Namecheap) is the GOAT

[OpenRouter @OpenRouter]

Introducing the Fusion API, the smartest compound model in the market.

Fusion achieves Fable-level intelligence at half the price.

How it works 👇

[Daniel M. @danielmrdev]

"Loop engineering" is just discovering what GSD and other projects shipped in December 2025.

Every week now there's a hot take:
→ "Stop prompting agents, design loops"
→ "The 5 components of the agent loop"
→ "Run sub-agents with fresh context to defeat context rot"
→ "Separate your maker from your checker"

And every time, I open GSD's docs and there it is — December 2025.

The phase loop: Discuss → Plan → Execute → Verify → Ship.
Fresh-context subagents for each phase. External STATE.md as memory across sessions. Parallel execution waves. A verifier agent that checks the executor's work.

All there. Before "loop engineering" was a term. Before the Codex agent loop post. Before Addy Osmani's article.

OpenAI "unrolls" the Codex loop → GSD had it running.
Addy coins "loop engineering" → GSD was shipping with it.
Boris says "I write loops, not prompts" → GSD users already knew.

The hype isn't wrong — loops are the right model. But the pattern isn't new. It shipped on npm in December. 64k stars later, the rest of the industry is catching up.

GSD was solving context rot before most people knew they had it.

[Daniel M. @danielmrdev]

@nachog Hola! Estoy interesado en el puesto! Pero me ha pillado de vacaciones (y sin portátil ni nada, solo móvil)… Puedo mandarte DM?

[Daniel M. @danielmrdev]

@0xSero For UI changes you can ask Codex to use Claude as a sub-agent to have Claude handle frontend/design work.

Just say “Use claude -p with an excellent, well-scoped, but un-opinionated (UI/UX-wise) prompt anytime you need a design change).”

I have a dedicated skill to do that.

[Xiaomi MiMo @XiaomiMiMo]

🚀 Better inference efficiency, lower costs, broader access.

MiMo-V2.5 Series API pricing is now permanently reduced — by up to 99% compared to previous pricing.
✨ Unified pricing across all context lengths.
MiMo Token Plans have also been upgraded:
• 5–8× more usable tokens at the same price
• Simpler and more transparent billing rules
🎁 As a thank-you to current users, all current Token Plan credits will be fully reset.
🎧 MiMo-V2.5-TTS remains free for a limited time.
⏰ Effective May 26 at 6:00 PM PDT.

These improvements are powered by continued inference optimization and serving efficiency upgrades across the MiMo stack.
🛠️ We’ll also publish a detailed technical blog on the inference optimizations later — stay tuned.

[DeepSeek @deepseek_ai]

We are making our discount permanent! 🎉

Enjoy building with DeepSeek-V4-Pro and bring your innovative ideas to life! 🚀

[Uncle Bob Martin @unclebobmartin]

I am absolutely more productive using agents. I don't know the factor but it's large. However much of that productivity is spent tuning the agents and hardening the product. I'm guessing 30%-40%.

Some might consider that a waste; but I don't. The software I'm creating nowadays is vastly more robust than I'd ever been able to create manually.

I don't mean that the code is better. I mean the surrounding tests are vastly better. I have a higher degree of confidence than I ever had manually -- even when I used very disciplined TDD and Acceptance testing.

And then there's the ability to quickly reorganize the modules and the architecture while keeping those robust tests running. That is a tremendous boon.

[Daniel M. @danielmrdev]

Three lines of Rust control how much reach your next post gets:

(1.0 - floor) × decay^position + floor

It's the Author Diversity Decay function. Your 2nd post in 4h loses ~50% score. Your 3rd loses ~75%. At a certain point your posts stop competing entirely.

Post less. Make each one count.

Full algorithm breakdown 👇
https://t.co/DenqCcYap5

[Daniel M. @danielmrdev]

The X algorithm is 207 files of Rust and Python.

You know what's NOT in any of them?

A penalty for posting links. Or hashtags. Or mentions.

The model receives exactly 4 booleans about your post: has_media, is_retweet, is_quote, is_reply. That's it.

What it penalizes instead is what I found most surprising.

https://t.co/DenqCcYIeD

[Daniel M. @danielmrdev]

I spent the weekend reading 207 files of xAI's algorithm code.

One thing stopped me cold:

There's a literal Kafka topic that decides if your post gets evaluated or ignored. If you don't get traction in the first minutes, the algorithm never looks at your post for quality. Never generates an embedding. Never shows it to non-followers.

Your post has minutes, not hours.

Full breakdown of what I found 👇

https://t.co/DenqCcYap5

[Daniel M. @danielmrdev]

https://t.co/y9nWuMV0T3

[Daniel M. @danielmrdev]

@trashpandaemoji @badlogicgames @opencode Yes, great deal. Flash has proven very capable to me. For complex planning, I switch to GPT.

[Daniel M. @danielmrdev]

👀
#BuildInPublic

[Daniel M. @danielmrdev]

@tibo_maker SEO

[Daniel M. @danielmrdev]

@juanmacias @juan_miqueo Me meto si ser invitado, Kimi K2.6 nivel Claude para mi. Este mes cancelo ya Claude Max 90€ y voy full con Kimi k
K2.6 por 39$… quizás deje Claude o Codex de 20€ como backup.