Want to build the tools developers and AI agents use to understand massive codebases? We're hiring an IC2 engineer on the Code Understanding team at @sourcegraph. Full-stack, remote, above-market pay. Come ship with us.
@Sourcegraph 6.12 dropped today, with an incredible lineup of new features. Check out the amazing new changelog with featured posts that absolutely fill the page.
The most difficult question to answer about code is *why* it was written the way it is. Being able to capture these conversations with agentic coding assistants, then subsequently index and search them could be a huge boost to understanding *why*.
Introducing tin, an open-source VCS for AI coding conversations, and my holiday hack project https://t.co/2oAfANTn6L
Today, all real design work happens in conversations with AI agents, and those conversations disappear into local caches. tin turns those threads into first-class, version-controlled artifacts, permanently linked to the code they produced.
tin retains the human thinking behind the code, the history of decisions that led to the code, and the provenance of the code. It provides a structural linkage between agent conversations and code changes.
@Sourcegraph@AmpCode This is personally exciting for me as a Sourcegraph team mate, but also invigorating for both @Sourcegraph and @AmpCode who will now have a sharper focus on building their awesome tools.
Big news today!
@Sourcegraph and Amp are now separate companies.
This move gives both products the clarity and focus they need to move quickly and serve customers better.
Proud of everything we are building!
Full blog in the comments, with why we’re making this move.👇
Just dropping today: @Sourcegraph and @AmpCode are becoming two different companies! This has been in the works for a bit now, and we are pretty excited for the opportunities this will open up on both fronts.
We are excited to announce today that Sourcegraph and Amp are becoming separate companies. I will become @sourcegraph’s new CEO.
I joined Sourcegraph a decade ago to help devs deeply understand the largest, most complex codebases in the world. With agents writing 100x more code and running 1000x more searches than humans ever did, the need for this is greater than ever.
Our users today are agents as much as humans, but they face the exact same challenges working in big codebases. If I update this API who is affected? How do I call this library? Does this functionality already exist? Have I updated all of the calls to this vulnerable code, or just the ones I know about?
In a large enough codebase, without universal code context, human or agent, you’re blind. If you’ve never worked in a 200+ dev org, you just won’t get it.
We believe:
1. Agentic search is the future of code understanding (today, our Deep Search)
2. Every coding agent needs universal code context (today, from our MCP+CLI)
3. The “outer loop” of the software development lifecycle *will* be automated by AI agents with full code understanding
We flipped the switch: Amp Tab is now on by default for new installs.
It’s our free completion engine for manual edits in VS Code. It’s fast, knows your recent changes and compiler errors, and suggests cross-file updates when needed.
Already on Amp? Run `Enable Amp Tab` from cmd/ctrl+shift+p.