Helio (@ItsOnlyHelio) header
Helio (@ItsOnlyHelio) profile picture

With a great life comes great responsibility to enjoy it.

Joined March 2019
Photo by Helio (@ItsOnlyHelio)Photo by Helio (@ItsOnlyHelio)Photo by Helio (@ItsOnlyHelio)Photo by Helio (@ItsOnlyHelio)Photo by Helio (@ItsOnlyHelio)Photo by Helio (@ItsOnlyHelio)Photo by Helio (@ItsOnlyHelio)Photo by Helio (@ItsOnlyHelio)Photo by Helio (@ItsOnlyHelio)Photo by Helio (@ItsOnlyHelio)
Launched our documentation ahead of our verity.md formal proof mechanism. From our testing, this is finding hard bugs consistently. I'm pleased about it but I'm hoping to get feedback from our users. Model-based testing will only trigger once we rank code as being buggy, or having many changes over time, or high criticality. Then it starts the flow. Now, every time your agent pushes code to Github, we run 3 threads in parallel to find as many security issues and bugs as we can. I know of no other pre-push reviewer in the market as strong as this. I am a firm believer that we should NOT be filling our repos with PRs that end up being stalled. We need proper code coming out of our agents. This is the main driver of Verity: to provide a claude-like or codex-like review just before you push. verity.md/docs/formal
1
2
233
1. build small Lean models of risky code 2. find sequences that break those models 3. run those sequences in your code 4. find bugs We're adding a new feature to Verity.md that runs this process automatically when you push code or on demand. We were inspired by @bcherny use of formal verification to build a simple way to find more bugs. The way it works is simple. It identifies the risky parts of your code, whether those are critical areas like payments or code that has high churn, and proposes areas of testing. Then Verity builds a small model in Lean of that code and tries many sequences to see if it breaks. Then if any is interesting, it replays those sequences in your actual code to confirm that there is a bug. if there is a bug, it will inform your coding agent. We've also built a skill to do this on demand. This is not formal software verification. This is using small models of risky parts of your product, blasting inputs and finding more bugs for your claude code to fix. This feature is being launched next week in Verity, which is completely free today while in beta. I'd love your feedback
3
1
7
308
RT @jaimefjorge: One of the reasons we built verity.md: because we need a knowledge base compatible with any coding agent or harness that g…
This quoted post is unavailable.
Just deployed gpt-6-luna and gpt-6-sol in Verity for agent-mode reviews. Every time your claude code works, verity automatically reviews with gpt-6-luna and sol. verity.md/
1
1
1
168
Mil comunicados valem mais do que as imagens.
Eles dizem que não há toque no Inácio. Que é um escândalo! O Manto Verde está de volta. Manipulação da verdade, Hipócritas, Avençados ou Idiotas Úteis? É escolher. #Sporting
1
1
314
É comparar a exibição do Porto até agora com a nossa de ontem. Comparar as decisões de arbitragem de um e do outro. E depois comparar o que se disse de um e o que se vai dizer do outro.
31
21
1
175
7,299
Os critérios do VAR de hoje e os critérios do VAR do jogo de ontem..... O futebol português é um risota.
26
76
1
861
26,310
Helio retweeted
Chief of Engineering at @EvolvTechnology on our AI Inventory. Learn more about it here eu1.hubs.ly/H0xdrC20
2
3
142
Acabei de assistir He-Man no Prime. Vi muita gente dizendo que era horroroso, e na verdade entendi o porquê. O filme é um tributo ao desenho original, do início ao fim, e ser tosco faz parte disso. A galera mais nova nunca iria entender a referência, mas quem assistiu ao desenho na infância, vai captar as reações exageradas, o Esqueleto canastrão (aliás, o personagem ficou absolutamente fiel ao original), e até a cena de risada tosca no fim. Tudo retratando e até tirando sarro do original. Não pagaria pra ver no cinema, mas valeu a pena pela nostalgia.
132
104
21
2,575
112,584
Steal this AI maturity scale to guide your team toward loop engineering. 👇 (sound on 🔊) Most teams think they're at one "stage" of AI adoption. They're not. They're at all of them at once. We spent the last few quarters talking to hundreds of software teams about how they really use AI and that surprised us. Most people picture AI adoption as a ladder: autocomplete → prompting agents → "loop engineering," where agents run scheduled tasks, wired into your systems, operating on their own. But the teams we talk to are doing all three at once. Autocomplete in one repo. Agents shipping features in another. Already running loops against their own infra. What's consistent is the direction: more autonomy for agents, hand off the recurring and mechanical work, let people move up to the part that still needs a human. Most teams are further down that road than they realize. More autonomy brings more entropy and runaway cost if you scale agents uncontrollably (we've watched companies do exactly that). Plus real change, like reworking your Git flows to handle looping agents. I pulled the best insights from those conversations into the video. Watch it and you'll probably spot where your team sits.
4
7
14
877
O Sporting não fala de arbitragens nem entra por esses caminhos, é legítimo, por isso teremos de falar nós. Aquilo que andam a fazer ao Zicky nas competições internas é vergonhoso. É a roçar o desumano. Estamos a falar dum jogador que é super correcto dentro de quadra, que não cria problemas com ninguém e respeita todos os seus adversários e marcadores directos, não há ninguém que não goste dele, e que continua a ser perseguido pelos árbitros como se fosse um rufia ou um jogador maldoso (é o total contraste disto). Neste momento, pode-se fazer tudo quanto se quer a Zicky. Agarrar, puxar, andar às cavalitas, empurrar, tudo. Não são averbadas quaisquer faltas, por mais evidentes que sejam. E pior ainda, faltas averbadas só ao jogador em si. Nestes últimos 3 jogos, se Zicky teve 1 falta assinalada sobre si foi muito. E ainda ontem “sofreu” 0 faltas, “cometeu” 3 faltas e até levou amarelo. E depois temos de ouvir os comentários sonsos dos pseudo-entendidos do Canal 11 que “Zicky faz muita falta ao futsal português” quando o pivot está lesionado e a ser gerido, quando nem se protege o jogador ou os atletas da sua posição. A continuarem a fazer tudo quanto querem sobre ele sem sanções, nenhum jogador dura muito tempo. Por fim, outros critérios e outras decisões. O primeiro golo do Braga no jogo 2 vem duma falta sobre Merlim no início da jogada, árbitros foram ao Video e disseram que não havia nada. Ontem, o segundo golo do Braga, à vista desarmada, não parece falta. Mas na repetição há um empurrão claro de Luís Fernandes nas costas de Wesley para ganhar posição (“não há qualquer infração”). Os mesmos árbitros que conseguiram vislumbrar uma falta de Wesley no seu golo anulado por uma “abordagem negligente” do jogador leonino, quando a agressividade média do Braga nos 3 jogos foi superior. Que as Finais tenham melhor arbitragem e que haja a coragem que não houve no ano passado, sobretudo no jogo 4.
Lugar garantido na 𝐅𝐈𝐍𝐀𝐋 da #LigaPlacard 🟢⚪️ Bravo, Leões 💚 #FutsalSCP
17
35
519
28,922
🌋 𝐎𝐁𝐉𝐄𝐂𝐓𝐈𝐕𝐎: Pintar o Jamor de 🟢⚪️ #TaçaDePortugal
58
1,086
68
5,241
215,319
Helio retweeted
Tudo aquilo que eles reclamam que beneficia o Sporting já os beneficiou esta época.
The state of the refereeing in Portugal is so bad. How on earth can you give this as offside?!
11
96
4,584
Helio retweeted
Now that we've 100%ed code analysis, we are proud to embark on a new journey. Today, we’re applying our linting and scanning expertise to the tangible world. Introducing Codacy Lint ‘n’ Scan Hardware: A curated line of premium linters, scanners, and detection tools designed to solve real-world problems with the same precision we brought to your codebase. Check out our new collection: shop.codacy.com/
2
1
163
Comunicado #SportingCP sobre as sucessivas ações que o FC Porto tem vindo a protagonizar nos últimos tempos. 👉 bit.ly/4tgctki
177
406
55
2,508
212,080
⚙️ VS Code started as a 10-person team building an editor. Ten years later, it’s used by millions of developers, supports 90K+ extensions, and has become a foundation many AI tools build on. In this behind-the-scenes conversation with @codacy, we talk about the early pivots between web and desktop, the bet on open source, why performance and core editing still matter as much as AI, and how agents are starting to shape what comes next. 📖 Read the full interview: blog.codacy.com/how-vs-code-…
7
16
2
116
13,001
The biggest takeaways/nuggets from my interview with @GeoffreyHuntley on AI-native software engineering and the Ralph loop: 1. Software development and software engineering are now two different professions, and one of them is over. Software development, the work of translating tickets into code, can now be done by anyone for $10-42/hour while they sleep. Software engineering, architecture, security, requirements breakdown, understanding failure modes, is where humans still matter. If you identify as a "software developer," you're competing against a bash loop. If you identify as a "software engineer," your job is to orchestrate the loops. 2. The moat you think protects your software product doesn't exist anymore. Geoffrey argues you can clone any SaaS product, even those with BSL licenses or proprietary enterprise code, using AI. He ran Ralph in reverse on HashiCorp Nomad's source code to generate clean-room specifications. When he hit gaps from missing enterprise features, he ran Ralph over their marketing materials and product docs to fill them in. Any company relying on licensing or code secrecy as a competitive moat needs to rethink their strategy. 3. Cursor, Windsurf, and every other AI coding tool are essentially the same thing: a loop that automatically copies and pastes. Geoffrey built these tools professionally and says the harness does almost nothing; the model does all the work. There's no real moat in the harness business when you're reselling tokens. The only differentiator is taste and UX. Stop evaluating tools and start learning the underlying patterns. 4. Ralph is not a product. It's an orchestrator pattern for running thousands of AI loops. The simplest version is a bash loop that deterministically allocates memory, lets the LLM pick one task, executes it, then starts fresh. The key insight: every loop gets a brand new context window. You avoid compaction (where the AI gets dumber as context fills up) by never letting the context window accumulate competing goals. Your institutional knowledge lives in specification files, not in the context window. 5. Specifications are the new source code. Geoffrey's workflow: spend 30 minutes in conversation with AI, drilling into requirements, making engineering decisions, building up specs. Then throw those specs to Ralph and get weeks worth of work in hours. The specs act as a "pin" that reframes every fresh loop with your domain knowledge. He doesn't hand-write specs. He code-generates them through structured conversation. Prototypes are now free. Refactoring is cheap. 6. The entry-level path into software engineering is closing fast. Geoffrey's company stopped hiring juniors for a year until they figured out how to interview for AI-native skills. There's already a cohort of juniors who've been practicing these techniques for six months. They'll work at a quarter of senior wages and outship them. If you're just picking up these tools today, you're behind. The new interview question: can you explain how to build a coding agent on a whiteboard? 7. Senior engineers who refuse to adapt are in more danger than juniors who embrace it. Geoffrey sees respected engineers taking hardline stances against AI ("it's installing fascism in your codebase"). Meanwhile, leadership teams are discovering Ralph and realizing three people can run the output of an entire org. When commit velocity and product velocity diverge that dramatically between adopters and non-adopters, founders notice. The hard line is coming. 8. AI is an amplifier of operator skill, not a replacement for it. If you're great at security and you get good at AI, you become a weapon. If you're mediocre and you use AI, you're still mediocre, just faster. The skill gap comes from "discoveries": learning the tricks, the loop-backs, the ways to close the automation loop. These techniques don't have standardized language yet. We're inventing the terms for the new computer every day. 9. Open source may no longer make sense for most use cases. Geoffrey, a former prominent open source maintainer whose land was funded by Open Collective, no longer uses open source libraries. His reasoning: every dependency injects a human into the loop. If there's a bug, you open a PR, chase a maintainer, wait. That's not automation. Instead, code-generate what you need. The exception: don't generate cryptography or security-critical code unless you have the domain expertise to verify it. 10. Programming languages now have a tier list based on how well AI agents can work with them. S-tier: Rust, TypeScript (especially with Effect.js), Python with Pydantic. These are source-based with strong type systems that reject invalid generations and work well with ripgrep for code discovery. F-tier: Java and .NET. Their DLL-based dependency systems don't work natively with the search tools AI agents use. The tradeoff with Rust: compilation is slow, so bad generations cost more time. 11. Corporate AI transformation programs are dangerously slow. Three-to-four-year rollouts with coaches and committees won't cut it when three founders in Bali can Ralph your entire product and undercut your pricing by 99%. Smaller teams ship faster. By the time the transformation is done, the market has moved. Geoffrey calls this the "Titanic moment": the boat is full, get the next boat. 12. We have a new computer, and that's why the legends are coming out of retirement. The last 40 years of computing decisions were designed for humans: TTYs, environment variables, slow language evolution to avoid breaking mental models. Now we have robots. What's the bare minimum a robot needs? Geoffrey sees this as the most exciting time in computing. If you're not excited about what you can now build, you haven't truly picked up the new computer yet.
35
110
35
714
64,429
Bem, se o Santa Clara nos oferecesse golos destes, nós não tínhamos de gastar dinheiro nos árbitros!!
Samu recebeu uma prenda de Natal atrasada.
14
24
3
616
28,389