Watch this and learn ... 1. How to use tmux to get better results from your agent 2. How to use an o3 tool call to "ask a senior dev" 3. How to structure your prompts so *you* can read them easier 4. In the future, are all pull requests going to have a link to the agent

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

This is super fast and costs $0 If you're coding in VS Code or Cursor, this one is for you

Amp now supports language-specific coding rules. This is nice for big monorepos, and it works even if you open the Amp CLI/extension in a subtree.

I'll be in NYC for the AI Engineer Code Summit (hosted by swyx ) in November. Would love to collab & chop it up with anyone who is also attending!

Sonnet 4.5 is now the primary model in Amp. I used it a lot this weekend while coding, and we saw incremental yet meaningful improvements that make it an obvious upgrade. Based on our eval/integration work and my personal experience: • It Just Works; keep doing what you were

GPT-5 is now the oracle in Amp

It’s official: the Amp TypeScript SDK is here. I’ve been working on it for the last few weeks now, and it feels amazing to share it with everyone.

Amp Tab suggestions come in streaks. Now you can actually see them stream in. The footer updates live ("1 more suggestion" → "22 more suggestions") as new ones load behind the scenes. Try it: "amp.tab.experimental.showHotStreakProgress": true

👀

We made Amp Free. It's powered by great tokens and tasteful ads. Agentic coding is now free for everyone.

Meet Amp's newest tool: The Librarian. Your cross-repo search assistant. The Librarian makes it easy to get clear and comprehensive answers to questions spanning beyond your current repo.

Today I ripped out /compact from Amp. I think we found something better: /handoff Here is me and Quinn Slack talking about why

We shipped a lot of stuff at Amp this week: • The Librarian subagent • Faster, cheaper search agent • Edit/Restore/Fork in CLI • Improved toolboxes • Handoff! • Amp Free without using training data So I decided to record a short walkthrough of it all

Talked with Tim Lucas about Amp Free: • Why we removed training • Can you use a combo of `free` and `smart` mode? • How is this different from free models w/other coding agents?

Amp can now read other Amp threads! Drop in the URL of a thread or @-mention it and Amp will then fetch the thread and extract relevant information. I'm very, very excited about this. Imagine how many Markdown files we no longer need!

More for less: the average paying Amp user now pays ~21% less per day vs. a few weeks ago, while # {threads,messages} {,per user}, users, & total revenue are all growing. We did this while keeping Amp's `smart` mode quality the same or better, and without subsidizing you (which

We're all learning a new way to code, with agents. Now you can see how other people actually use them.

We’re making Sourcegraph and Amp two independent companies. Sourcegraph will be totally focused on code search and code understanding for big codebases, in a world where AI is writing and searching 1000x more code. Amp is becoming Amp Inc., an independent research lab,

Any application supporting React Server Components is potentially vulnerable. Are you? sourcegraph.com/blog/find-affe…