One of the most cracked teams out there!

I'm excited to share Github Sage, a new way for developers to search, get support for, and build off open source software. loom.com/share/222b4f4c…

There’s so much OSS out there, and I’m like a kid in the candy store. I want to try it all, but there are only so many hours in the day. Sometimes I just want to ask some high-level questions before deciding whether to invest in a repo. We made Code Sage to scratch that itch:

𝐏𝐫𝐨𝐮𝐝 𝐭𝐨 𝐟𝐢𝐧𝐚𝐥𝐥𝐲 𝐚𝐧𝐧𝐨𝐮𝐧𝐜𝐞 𝐨𝐮𝐫 𝐟𝐮𝐧𝐝𝐫𝐚𝐢𝐬𝐞 𝐟𝐨𝐫 𝐀𝐠𝐞𝐧𝐜𝐲 We’ve secured $2.6M to build generationally important technology and bring AI agents to life Here’s a look at future we’re building with agen.cy (🧵):

Exciting update: Storia AI Sage now allows you to index your own codebase, in a fully self-serve fashion. All you need is a public Github link, and 3 minutes later you'll have an AI expert assistant that can answer questions and contextually generate code related to your repo.

Google truly mogging everyone out there!

I'm hiring in Bangalore. If you want to be part of this fast-growing team and help shape the future of AI Agents, apply here: lnkd.in/geFNC8NN. Please like, share, or tag anyone who might be interested—we’d love to hear from you!

This is true but somewhat alleviated if you spend time debugging/going deep into the results/process with the ICs

Do folks know of a company/product that integrates with my chat (Slack/Google Chat), indexes/searches all channels, and composes answers based on chat history?

One of my fav authors has a new blog out!

Windsurf's lawyers in the last few days

So much of the AI vibe app building discourse is about Lovable vs Replit, but in my experience, typically Bolt is both cheaper and gives me better results consistently.

Startup Truths by Quinn from sourcegraph. - Everything is changing. - Cursor is already previous generation.

With all this cloud and Anthropic hype for coding, a counter datapoint: I recently tried Gemini 2.5 pro in cursor agent mode and found it both faster and more to the point compared to Anthropic models.

My experience of using GPT-5 for 24 hours - it's worse than Sonnet 4 on every single coding task I tried. Worse for debugging, worse for design, and worse for overall code generation. Even though it's free in Cursor for this week, will be switching back to Sonnet 4.

It seems that the latest claude code version (1.0.80) is broken. Installing 1.0.79 worked for me.

Trying out Figma Make for building a simple website. Missing a ton of features compared to Replit, Bolt, and Lovable; but the simple, clean visual output is unparalleled. Figma's probably running away with the vibecoding market.