Augment Code

Hello, world! We’re Augment Code, the first developer AI purpose-built for teams. What makes Augment different? Context. Every Augment feature is context-aware. Read more: https://bit.ly/4fdJ1UU

See Augment live in action — June 11, 12:30pm ET. Our AI understands your codebase. In 30 mins, we’ll demo how our features, Agents, and Remote Agents can help developers: ✅ Ramp up fast ✅ Write better unit tests ✅ Refactor legacy code Live demo + Q&A. 👉 Save your seat: https://lnkd.in/gvW4yVNw

Big news.

Big news today: Remote Agents: now live for all Augment Code VS Code users. CEO Scott Dietzen told Frederic Lardinois of The New Stack: “The thing that made our original agent unique was this deep knowledge of customer code,” he explained. “In general, agents often get lost. But if you can have an agent that’s an expert in your code base rather than a novice, they can be entrusted to do a great deal more. And so this context engine that our AI research team built over two and a half years turns out to be a profound differentiator for agents, because the more autonomous you’re looking something to be, the more you critically rely on that context.”

Kicking off a search for an experienced Technical Sourcer to join us at Augment Code! We’re assembling a deeply technical team to solve one of the hardest problems in AI: making software development radically more productive without taking developers out of the loop. This role goes well beyond owning an outbound motion. We’re looking for someone who: – Knows how to build trust with world-class engineers – Thinks in systems, and gets joy from fine-tuning a pipeline – Can context-switch without dropping threads, and knows when to push and when to pause – Balances rigor with speed, and cares as much about the team’s success as their own – Is deeply proud of their work product We're scrappy and we move very fast. The bar is high. If that sounds like your kind of challenge, we'd love to hear from you. https://lnkd.in/esWVyfUV

In San Francisco this week for AI Engineer World’s Fair! If you’re in town for workshop day, Forrest Brazeal and I are running a workshop with Augment Code’s Matt Ball on Navigating deep context in legacy codebases. In this session we take on the source code for the guidance computer that landed the 1969 Apollo 11 astronauts on the moon—written in assembly! Leveraging cutting-edge AI tools to navigate and understand the codebase, we then turn legacy code into something any modern app developer will recognize (and see if we can use it to land the lunar module—something you might not want to trust to vibes). This was a LOT of fun to put together (and to source prizes for), and I hope to see some of y’all at 3:30 today! https://www.ai.engineer

🆕 Mermaid diagrams in Augment Agent Ask: 💭“Show the data flow for Foxglove MCAP generation in our system” 💭“Show me a diagram of my data model.” 💭“Visualize the request flow in onUserLogin.” Agent reads your codebase, understands structure, and renders a Mermaid diagram—directly from your prompt. No setup. No config. Just type.

🚀 After a week of vibe coding with Augment Code (IntelliJ plugin powered by Claude Sonnet 4), I’m thrilled. As a lifelong IntelliJ IDEA user, it hurt to open Cursor (a VS Code fork) just to work with an AI agent. My workflow looked like this: 🎧 Vibe coding in Cursor 💻 Real coding in IntelliJ IDEA 😩 One project, two IDEs — not exactly a seamless experience I had tried other IntelliJ plugins — like Junie — but none of them came close to the AI experience Cursor offered. But now? Augment Code finally brings that Cursor-level AI "magic" into IntelliJ IDEA. If you live in IDEA like I do — give Augment Code a spin. 🙌

Prompt length isn't your enemy. Missing context is. If your agent consistently fails, it's usually because it wasn't given enough to succeed — not because the prompt was too long.

Forget vibe coding. Seriously. My Augment Code team was able to rigorously test Anthropic’s Claude Sonnet 4 pre-launch. This version is the G.O.A.T. (Greatest of all tested). With Claude Sonnet 4, engineers can rest easy knowing it does the work and checks it. - Edits are focused, surgical and relevant to coding in production. - Complex tasks are handled - with exciting potential for Remote Agents. - Noticeable improvements from 3.7, especially for Agentic workflows and staying on task. Already have your code base hooked up to Augment Code? Claude 4 is rolling out now. New to the game? Sign up for a 14-day free trial and see the power of Augment's Context Engine with Claude 4. https://lnkd.in/g3TiqGzi Happy coding!