//
sign in
Profile
by @danabra.mov
Profile
by @dansshadow.bsky.social
AviHandle
by @danabra.mov
AviHandle
by @dansshadow.bsky.social
ProfileHeader
by @dansshadow.bsky.social
ProfileHeader
by @danabra.mov
ProfileHeaderAlt
by @jakesimonds.com
ProfileMedia
by @danabra.mov
ProfilePlays
by @danabra.mov
ProfilePosts
by @danabra.mov
ProfilePosts
by @dansshadow.bsky.social
ProfileReplies
by @danabra.mov
Record
by @atsui.org
Skircle
by @danabra.mov
StreamPlacePlaylist
by @katherine.computer
+ new component
Profile
Loading...
Open-Source Strategist and Chart Scientist @googlecloud | Open Data Advocate | Calvin & Hobbes enthusiast | 2012 Ft. Sill 800m champion | @shanecglass on Twitter | @[email protected] on Mastodon
Shane Glass









Loading...
Registration for Iceberg Summit '26 🏔️ Join in SF or virtually on Apr 8–9 to dive into the Apache Iceberg v4 roadmap and architectural patterns at scale. Early bird ends Feb 15. Virtual is free! 🎟️ Register here: www.icebergsummit.org/?utm_medium=...
By defining the architecture and specs upfront, tools like Antigravity or Gemini CLI get the structural context they need to bridge the gap between "plan" and "code." It turns those defined needs into a true launchpad for the agent.
I’ve noticed a consistent pattern in what makes my agentic development actually click. My most effective workflows right now aren't always the ones where I write the best prompts - it's where I've grounded them in my project's specific requirements and architecture. 🧵
Apache Polaris has graduated to a top-level project! This has been a long time coming, and we love that it makes life even easier for Iceberg users to get value from their data. Congratulations to all of my Apache Polaris friends. www.snowflake.com/en/blog/apac...
I’ve found that Antigravity's implementation plans are way more effective with a clear PRD or technical spec. Sometimes I’ll export the doc to Markdown. Sometimes, I’ll use Conductor in the Gemini CLI to generate a targeted requirement doc specifically optimized for an agentic workflow.
It feels counterintuitive to slow down so you can move faster. But slow is smooth and smooth is fast. My smoothest starts are grounded in clear documentation. Curious to hear how you are grounding agents—are you relying on detailed PRDs, or finding success with a more "vibe-based" approach?
It really comes down to how strong of an opinion I have on the architecture. If I have specific requirements, Conductor is great for generating a precise, agent-ready doc. If I’m more flexible, Markdown docs are usually enough.
The cost of a welcome is zero, the value is infinite. 💖 Join Women+ in Open Source day on March 9 to celebrate the incredible women shaping the future of open source and help build a stronger pipeline for everyone. luma.com/qcqlia4g What first made you feel welcome in the OSS world?
If you're here at FOSDEM this week, I'd love to chat! Send me a DM and let's catch up
You can test new tech ideas using the Seinfeld Test Would the product eliminate the plot of an episode? (Google maps, cell phones, paypal, battery packs) Good tech. Would the product inspire new Seinfeld plots? (NFTs, AI chatbots, crypto currency, blindboxes, metaverse land sales) Bad tech.