Hello Seattle 👋

ugh, just wrapped up a health check. everything’s fine, but not optimal. now that i’ve got the details, it’s easier to pinpoint what to fix and track progress. tldr: time to make health a priority again.

personally invested in judging @nerdearla MX 2026 CFPs this year and honestly can't wait to see what ideas the community cooked up

first time watching @hahnbeelee speak and you know it's serious when they bring tweets to the keynote

ok y'all im panicking. I have a chat with the author my favorite book on communities in two weeks and idk how to react to this

One of the decisions we took when we made our talk about "Why tech became obsessed with taste", was to push the conversation about the relations of taste with power and capital.

connecting friends to great jobs opportunities is one of the most fulfilling things I can do

hear me out, evolving pets for Codex/ChatGPT that level up with token usage. throw in some variety and rewards for heavy users

Feeling so left out 😔

we've been testing GPT-5.6 Sol and Terra for coding agents and code review. Check our blog for all the details on this www.coderabbit.ai/blog/gpt-5-...

OpenAI GPT-5.6 Sol and Terra: Benchmark

OpenAI’s GPT-5.6 family includes capability tiers: Sol as the flagship model, Terra as the lower-cost option, and Luna as the fastest, lowest-cost tier.

coderabbit.ai

team really cooked on this one 🥕

really impressed by how well @cursor plugins work and how much you can build with them. loved being able to add hooks, rules, and other elements other agents don’t support, and it noticeably improved our plugin’s results.

So sad to see GitHub beginning to shut down Classroom. We used it heavily to help students learn on GitHub, and it helped scale assignment distribution and reviews for thousands of students.

favourite thing from the @Railway peace campaign is "deshrimping"

Sol, Terra, Luna

Sauna and cold plunge day in Sausalito

fixed the background music for the Fable 5 release video

playing around with a new project. wild how much you can build when you give it the right prompt and context

next time someone says there are no Latinos working/using AI, I'm showing them these photos

I spent time testing NVIDIA’s Nemotron 3 Ultra. In our benchmarks, it got close to baseline review quality. But it also needed a lot of retries but with a good harness and validation, it gets interesting fast.

I need an encyclical reading club with wine and charcuterie boards.

the luck of being in Mexico City the same week one of my favorite artists has a banger expo

Hey y'all we are organizing an AI meetup for Latin@s in SF next Thursday at the @coderabbitai office. @jeduan: What he builds with vibe coding after his workdays at Airbnb. @fforres: His journey from OpenAI to building a company between Chile and California.

Hawaii was the perfect place to go offline, touch grass, and unplug for a week.

usage of @coderabbit CLI since we have optimized it for agents

We’ve been improving our code review UI to make easier to navigate code changes: - Semantic Diffs added - Complexity analytics - Improved navigation Here’s what’s new:

100 likes and I make this my profile picture

tech marketing people who stopped learning in 2020 are getting left behind hard the ones who adapted are more valuable than ever

you know it's a quality event when you see @hopes_revenge at it.

watching @coderabbitai review and @claudeai fix the same PR live at the Code with Claude event this is wild