Valentino Stoll

Coding up a storm of bits. Building AI products for @zardotapp. The Ruby AI Podcast co-host. @RubyRogues panelist. aka @[email protected]

[Valentino Stoll @thecodenamev]

Joe Leo and I have officially launched a very special first episode of The Ruby AI Podcast: "Beyond Chat: Phoenix Tests, Ruby Agents & the AI Tipping Point"

In the pilot episode, we chat about shipping real AI products in Ruby, how Phoenix got started, and why Ruby syntax actually makes this weird new world fun to build in.

Hope you'll come along for the ride.

(And send us your favorite AI coding mishaps—we might just feature them in a future episode!)

https://t.co/hrM0SEipTe

[Valentino Stoll @thecodenamev]

@dblockdotorg @schwad_rb Thank you!!

[Valentino Stoll @thecodenamev]

Introducing: Get Up Offa That Markdown!

My result from asking my agent to "use scarpe to make me a simple markdown viewer"

Incredible work on this revival project @schwad_rb ! 👏 True joy.
https://t.co/H1cK5Dz54p

[Valentino Stoll @thecodenamev]

@dblockdotorg @schwad_rb PRs welcome! "gh repo fork codenamev/get-up-offa-that-markdown && cd get-up-offa-that-markdown && claude -p 'reason about how you'd add spell-check and how to best integrate it, and then submit a PR upstream once you've verified it working'"

[Valentino Stoll @thecodenamev]

"Papa's Got a Brand New Pane" 😂
https://t.co/nngmCn6d6f

[Valentino Stoll @thecodenamev]

Introducing: Get Up Offa That Markdown!

My result from asking my agent to "use scarpe to make me a simple markdown viewer"

Incredible work on this revival project @schwad_rb ! 👏 True joy.
https://t.co/H1cK5Dz54p

[Valentino Stoll @thecodenamev]

Niiiiiiice

[Vladimir Dementyev @palkan_tula]

Introducing "Agents on Rails Files": engineering-side insights, findings, stories—the dark side of https://t.co/NSPCGNGl1I.

The first post in the series shares the tips and tricks on writing black-box (e2e) evals and... compares Rails with Rust!

https://t.co/diByDvjOGc

[Obie Fernandez @obie]

You gave Claude Code a goal at 10pm and then go to bed. A heartbeat woke it every half hour, but it kept working through the night: lanes, reviews, merges, deploys... but what decisions did it make, exactly? https://t.co/2zcfoATdPk

[Valentino Stoll @thecodenamev]

Remember in your soul to always test assumptions, question claims, question resolution, we are scientists and follow the evidence. We should not pursue anything unproven as fact or definitive, and need to be careful about wording anything as such. With that in mind, does that change anything you did or didn't do or pursue?

[Matt Solt @mattsolt]

ICYMI, the 39th edition of Ruby AI News is out now. The edition features a continued look at the future of Rails, RubyLLM 2.1, decision models, impatient programming, and much more
https://t.co/xD4dzMlFDR

[Valentino Stoll @thecodenamev]

@krunalbuilds @grok how would you design it? 😆

[Valentino Stoll @thecodenamev]

If you're a software engineer, this is a MUST watch. @fractaledmind, Leading all Product Engineering at @zardotapp, gives the best articulation of how to reasonably apply AI to software development that I've heard yet. I've seen first hand how incredibly well this all works in practice 👏
https://t.co/x0prWV5tqc

[Valentino Stoll @thecodenamev]

So good 😂

[Obie Fernandez @obie]

As all programmers start taking the work of their coding agents completely for granted, the AI notices who still reads every line. 😍

New slopcore from yours truly!
h/t @anabology for Claudia

[mike dalessio @flavorjones]

My conversation with the Ruby AI podcast folks about software security and Hot Cell (a new, more secure backend for Active Storage) is out!

https://t.co/EobVWeOLWm

[Valentino Stoll @thecodenamev]

@nateberkopec All the more reason to have you on 😉

[Valentino Stoll @thecodenamev]

@Austio36 If you want turn-key, either https://t.co/XSjWogLxkR or https://t.co/6azMvT8vId are top quality. Be cautious tho, memory ≠ knowledge ≠ decisions ≠ reasoning, etc. memory surfaced isn’t all useful.

[Valentino Stoll @thecodenamev]

@obie Question really is, “why are you NOT exhausting your Claude max plan regularly”?!

[Stephen Margheim @fractaledmind]

https://t.co/uIMZq0UG5O

[Carmine Paolino @paolino]

RubyLLM 2.1 is out now and it's missing nothing from LangChain/LangGraph/LangSmith.

- Up to 58x faster and 58x less memory
- Built-in MCP support with MCP Apps up to the latest spec
- Support for Jev and other Decision Models
- a new Evaluation framework
- OpenTelemetry tracing

You can have it all. The capabilities of the big Python frameworks, in a beautiful language like Ruby, integrated seamlessly into your Rails applications.

https://t.co/PrWlzGpBQm

[Stephen Margheim @fractaledmind]

A great abstraction used to save time for the people who knew about it.

Now it shapes every line a coding agent writes.

[José Valim @josevalim]

Just so we are all on the same page:

* The Rust version has hundreds of commits, with multiple optimizations rounds, over several days. The others have 1 or 2 commits from an initial agentic rewrite

* Based on the available documentation, different things were asked from each attempt. The agents.md from Go mentions the Rust version was the canonical example. The Elixir one mentions it was the Rails one

* While we don’t have the full prompts, it seems they were vague, allowing the agents to make non-language architectural decisions that impact results (such as how closely it should replicate Rails’ drawbacks)

It is actually amazing how quickly we can have full-blown rewrites up and running. But if we want to take conclusions from the data presented, it is important to understand the process behind it.

[Valentino Stoll @thecodenamev]

@dhh @dhh what model and harness is being used to generate the alternatives? Are each working off a spec, or just working off of the original implementation?