All software as we know it is under an immense threat right now. Proud to be working at a company with so many talented people making the internet a safe place again. Kudos to our team & to our founders who have pledged $1M in credits for folks to upgrade their security.

Great vibes and even better tea 🫖 Here’s what I took away: • @romainhuet and the DevEx team shared their thinking on “What Comes After Loops?” One practical detail from the slides: keep a worklog of the goal, current instructions, who owns what, unfinished work and decisions. When instructions change, reconcile them with work already underway. After a context compaction or handoff, read the log and check the actual state before continuing. That gives an agent something concrete to pick up from when the work spans multiple sessions. • @dkundel made a great argument for defaulting to one agent session before building overcomplicated orchestration. Most tasks can be handled perfectly well in one Codex thread. Give it responsibility for the outcome and let it delegate independent work where that helps. The “Coordinate” and “Manager” skills shown on stage made this concrete: split work with clear ownership, let agents resolve dependencies directly, and name someone to assemble and check the result. If two agents are working together, define which parts each can edit and use cloud env so they don’t overwrite each other’s work locally. Also, avoid creating too many org layers and therefore managerial debt • @whoiskatrin talked about what it would take to keep Codex running for 10,000 hours straight. The “keep CI green” example made this tangible. The accompanying AGENTS.md told the agent to run the failing test before and after a fix, keep notes in ci-notes.md, and open a PR explaining what it tested. It also set clear boundaries: don’t disable tests to make CI pass, don’t trigger unnecessarily large CI runs, and leave the PR for review. Give it a way to check its work. Tell it what matters and when to ask for help.

This is a good day for open-source software! CodeRabbit has pledged $1M in AI Security Deep Scans to OSS and now we have most powerful models to help bullet-proof the foundational software of that powers the internet for the next wave of cyber attacks.

@coderabbitai CLI v0.9.0 is out New: cr code, to manage cloud agent tasks from your terminal. Messages, side questions, plans and delivery, all without leaving the shell Plus shell completion, richer review JSON and signed Linux releases cr update

CodeRabbit is a SpaceX Super Intelligence Launch Partner! @SpaceXSI Now in the Grok Build Plugin Marketplace: build with Grok Build and independently review changes with CodeRabbit. Read the announcement

Forgot to grab a good photo with you all, but I really enjoyed joining you at ThunderPlains @techlahoma! Thanks to @vlucas and the Techlahoma team for having me. Slides & Github Repo in the comments 👇

Who said Hamburg doesn’t have a startup scene? Founder dinner at Chefs Warehouse last night. Thanks @Bojan_Jukic @Maltissimo11 @LennartPrange @CedikZou @c_kroeber @andreas_wissel @najmuzzaman and everyone who came. Special thanks @AI_BEAVERS, Aleks and @vlady_nyz! 🦫

This one was such a blast! Thank you to all the judges and the killer orga team! 🙌 @AJs_AI @roangws @DaLucasGonzalez @shreypandya @Claude_Memory @allietheicon @utkarshg20_ @xrhuang10 @_nikhilhooda @SourabhMane_

4/6 🫡 - People in SF (and tech week) keep amazing me... Even with Karl (the fog) making the beach less inviting over 50+ people still showed up! We did yoga, surfing, shared conversations on Software, AI coding, tech or just life in overall. Epic Morning. Thanks everyone!

That's the paradigm shift you have to think about when working with a decision model like Jev. Don't make another AI call just because your code took a different branch. You have to ask more questions, all at once...

3/6 ☑️ 😅 - Another eventful night at @Intuit OpenSource - preaching the Rabbit and sharing the stage with the amazing @aparnadhinak from @arizeai talking about how to build self-improving agents...

My SF Tech Week calendar got a little out of hand... 😁 I’m co-hosting and speaking at six events across SF and Mountain View next week. Here’s where you’ll find me: 🏃 Monday, Oct 5 · 8 AM - Rabbits on the Run An all-paces run, then breakfast and a little rave at @coderabbitai HQ. Warm-up at 8, run at 8:15, breakfast beats at 9 on our rooftop. 🏭 Wednesday, Oct 7 · 5 PM - Factory Reset w/ @mastra Short, opinionated talks about the agentic software development lifecycle: AI code review, open source, autonomous engineering, SW Factories, and what’s actually working w/ @abhiaiyer ⚙️ Thursday, Oct 8 · 4:30–8 PM - AI & Platform Engineering Developer Meetup with @Intuit OSS I’m in Mountain View talking about scaling engineering judgment when code is basically free, alongside Intuit engineers and @arizeai’s @aparnadhinak Dhinakaran exploring agents that improve through feedback and evals. 🌊 Friday, Oct 9 · 8 AM - Waves & Breakfast with CodeRabbit Surf, a cold plunge, coffee, and breakfast at Ocean Beach. Beach people welcome, even if you’re skipping the surf. 🌇 Friday, Oct 9 · 6–9 PM - October Tech Rooftop Mixer Closing out my week with Founders & Friends and Founder Social Club: founders, builders, investors, and sunset conversations in SF. Join us 🔥 Saturday, Oct 10 · 4 PM · Menlo Park - Nar Nights A fireside chat with an actual fire, under the redwoods. A small circle of founders, investors, and researchers sharing stories, with pillows and carpets on the ground. If you're in the city and any of that sounds interesting to you, I'd love to meet you! See ya there! 👋 P.S. Links below

When an agent writes the code, who owns the result? @trmlabs Rahul Raina and CodeRabbit’s Harjot Gill discuss the test evidence, human judgment, and team habits a company needs to ship at AI speed and trust the outcomes. Watch the full video below!