We are a team of researchers and engineers at @GitHub who investigate the future of software development.
The Future
Joined July 2020
- Tweets709
- Following47
- Followers29.6K
- Likes1.3K
Pinned Tweet
How do you _guarantee_ your agent's code is correct? You prove it.
Introducing Proof Driven Development Kit (PDD Kit) from @MSFTResearch.
PDD Kit helps turn your plan into a formal spec, then proves the agent's code meets that spec for every input. No vibe verification. PDD Kit is integrated with @GitHubNext's Chopin, so you can plan as a team.
PDD Kit takes an app's login flow from intent to a verified pull request:
It discusses your code, infers your intent as it goes, and proposes a formal spec in Verus, a proof language for Rust.
Once you approve the spec, the Verus verifier automatically proves the code meets it.
Where proof isn't possible, it falls back to empirical verification with tests and computer-use agents (CUAs).
It works on Rust today. No new language, no change in how you produce the underlying code, making formal methods more accessible.
The goal is a verified software factory, where engineers review specifications instead of code.
Come see it in action at GitHub Universe, Oct 28–29. Swing by the Speakeasy and ask for GitHub Next!
GitHub Next retweeted
so satisfying
Built a skill to bring the magic of the OP-1 to agentic development and wrote a post about it on @GitHubNext (sound on for maximum clickage)
GitHub Next retweeted
Built a skill to bring the magic of the OP-1 to agentic development and wrote a post about it on @GitHubNext (sound on for maximum clickage)
GitHub Next retweeted
So much is moving towards computers becoming actual terminals again, running their compute *somewhere else*. Even the personal stuff.
The Hypercloud Movement™
GitHub Next retweeted
How do you solve the trust problem with AI-generated code?
AI agents can generate code at a scale too great for humans to review line-by-line. That makes it difficult to ensure the code will do what the developer intended, and it leaves the final output vulnerable to bugs and security risks.
Proof-Driven Development Kit turns plain-language developer intent into a specification in Verus, applying formal methods traditionally used for safety-critical systems. PDD Kit uses mathematical proofs to test agent-generated code against the spec where possible, and clearly indicates where the proofs can’t reach. The result is more reliable agentic code.
Researchers @MatthaiP39719 and @alexlavaee break down Proof-Driven Development in the MSR blog: msft.it/6019af2V7
GitHub Next retweeted
Making 50% of the world’s code verifiably secure.
Traditionally, formal methods have been relegated to verifying safety-critical systems like operating systems, hypervisors, databases, etc. We’re extending this approach to formally and empirically verify code correctness and security for full-stack web applications using Proof-Driven Development Kit (PDD Kit).
A large portion of the world’s software consists of web applications, and we see this as a starting point for our everyday software.
Check out our blog here: githubnext.com/projects/pdd-…
From @GitHubNext x @MSFTResearch
GitHub Next retweeted
We've just published an in-depth write-up about the PDD Kit - collaboration project between us (@GitHubNext) and @MSFTResearch focused on formal verification in the world of AI agents. 🤖
Check it out:
githubnext.com/projects/pdd-…
How we can secure 50% of the world's software.
Traditionally, formal methods have been mostly used for safety-critical systems like operating systems, hypervisors, and databases. We're extending this to formally and empirically verify code correctness and security for full-stack web apps, one of the most common types of applications*, using the Proof-Driven Development Kit (PDD Kit).
Coding agents now write code faster than anyone can review effectively. PDD Kit turns what the developer meant into a Verus specification. It then proves the AI-generated code meets that spec for every input.
Researchers @alexlavaee, @MatthaiP39719, @Mappletons, and @k_cieslak break down Proof-Driven Development x Chopin in the blog here: githubnext.com/projects/pdd-…
*44.4% of respondents work as full-stack (27.3%), back-end (13.5%), or front-end (3.6%) developers. Stack Overflow Developer Survey 2026, n = 25,767.
GitHub Agentic Workflows has graduated out of the lab and is now generally available! We're excited to see more projects and businesses leverage AW to automate the things that could not be automated before AI.
github.blog/changelog/2026-1…
Missed the closing keynote at GitHub Universe? You can see @Mappletons from GitHub Next demoing Chopin, our exploration of multiplayer planning for the agentic era
youtu.be/Gh-hg7lwA0o?si=Ys5q…
GitHub Next retweeted
ALT Maggie giving her talk
ALT GitHub Next cooking
ALT Salut Chopin
ALT Chopin in action
Furiously making last minute preparations for our talk tomorrow, here's a fun mixtape from the Next team to jam to during @githubuniverse!
mix.githubnext.com/
GitHub Next retweeted
Today is my first day at @GitHubNext!
GitHub has been one of my favorite places to hang since I signed up 18 years ago(!), a mere month after its founding.
Excited to join this talented team to try and figure out what comes next for our craft ...
🇯🇵 In Japan? Idan Gazit from @GitHubNext and @JasonEtco from the @github Copilot team will be speaking at @findy_tools AI Engineering Summit on December 14+15. ✨
／
「AI Engineering Summit Tokyo 2026」特設サイト公開✨
＼
AI時代の開発に挑む、すべてのエンジニアと技術リーダーのためのカンファレンス。 12月の開催が決定しました！
国内外から、AI開発の最前線を走る企業が集結します🌎
#AIE2026_findy
▼ 基調講演
・Chip Huyen氏（「AIエンジニアリング」著者） @chipro
・Idan Gazit氏（GitHub Head of GitHub Next）
・Jason Etcovitch氏（GitHub Copilot Agents, Principal Software Engineer） @JasonEtco
▼ 特別講演企業（一部抜粋）
・Duolingo
・Meta
・Uber Technologies Inc. @mykolaskrup
・Vercel @elsigh
・株式会社村田製作所
・日本電気株式会社
・東京海上日動システムズ株式会社
▼お申し込みはこちらから
ai-engineering-summit-tokyo.…
GitHub Next retweeted
Found time last week to cleanup the homepage & docs of Agentic Workflows, one of the very successful @GitHubNext projects of the last year. TLDR: run your own software factories on GitHub Actions
Previous page was vibed by agents and backend developers. Took me a day to redo it.
GitHub Next retweeted
Had a wonderful chat with @GergelyOrosz - always nice to go deep on ideas in a proper long chat. Got to show off some notebook sketches and design jigs with agents too.
What can us, engineers else learn from designers and design engineers? Turns out there’s lots, based on talking with one of the best design engineers in the industry, @Mappletons.
Timestamps:
00:00 Intro
03:24 From anthropology to tech
10:18 What does a designer do?
18:23 How Maggie works
24:55 The case for planning with physical tools
31:53 Why Maggie is learning woodworking
33:13 Design engineers and engineering constraints
38:49 How Maggie uses Figma
40:30 Design at GitHub Next
45:12 How has AI changed design
50:37 When models design and why humans are still needed
53:30 UX and UI
58:29 Capability gaslighting
1:00:33 One Developer, Two Dozen Agents, Zero Alignment
1:07:21 Craft and AI tells
1:14:17 Visual gardens, home-cooked software, and barefoot developers
1:21:02 Advice for engineers and lessons from anthropology
1:25:34 Book recommendation
Brought to you by:
• @turbopuffer – a vector and full-text search engine built on object storage. It’s fast, cheap, and extremely scalable. turbopuffer.com/pragmatic
• @AntithesisHQ – verify your system’s correctness without human review or traditional integration tests – and avoid bugs or outages. antithesis.com/pragmatic
• @EntireHQ – every agent prompt, tool call, stored in your repo, and mirrored. Set up with one click: entire.io/pragmatic
Hope you enjoy this unusually visual episode - thanks again, Maggie!
github.com/githubnext/localj…
Not everyone on the team has access to Jev yet. Spent a morning cobbling together a poor man's Jev on top of omlx for local use. Benchmarked and eval'ed a variety of models including diffusiongemma and a variety of autoregressive models (Qwen MoE, Gemma 4 MoE, and Gemma 4 e4b/e2b.) Benchmark report is in the repo.
This is 100% promptcoding but hey the evals look okay, speeds are pretty good on local machine (m5 max 64gb), and the LocalJev server exposes an API that can be used with the normal Jev API wrapper libraries.
Messing around with @typesafeai Jev for real-time styling
🔤 🔜 🎨
If you can define a bounded, semantically meaningful property space, you can project a natural language expression into that space in one generation step
From our friends on the Applied Science team, great stuff!
We're finally starting to see the multimodel future with an intelligent routing layer. As time goes on, we _should_ care less about the models, because A) they're all "good enough" and B) AI should be optimizing routing.
Plus, imagine a world where we do some inference on local hardware. Coding is certainly sensitive to parameter count, but a lot of the coding workload is not about code generation, but reasoning about English/NL. That part of the agentic workload doesn't necessarily need the flagship models, or even the cheaper cloud models. Those subagents don't necessarily require long context windows (which in turn requires massive amounts of RAM and memory bandwidth). Companies are buying powerful machines for their engineers — machines that can do some of this work, but only if there's a routing layer that can route to them. Excited to see the first steps in this direction!
4.9 percentage points higher verified task quality. 67% lower estimated cost.
Project HydraFusion delivered those results against Claude Opus 5 on Terminal-Bench 2.1 in controlled offline evaluations.
HydraFusion orchestrates the models and workflow for each coding task. You select it like any other model. ⬇️
github.blog/ai-and-ml/github…
4
18
4,866