@OriSilver Appreciate you open-sourcing this, Ori. Progressive disclosure, artifact-first output, and explicit unknowns are genuinely useful.
My reservations: heuristic scores risk false precision, platform claims need evidence/evals, and Markdown instructions still need security review.
Nod (@hellonodapp) is expanding from a personal AI coworker into a shared intelligence layer for teams.
We’re onboarding a small group of design partners: shared context, work-tool integrations, custom integrations, SSO and enterprise security support.
DM me or apply: https://t.co/nkSIf15Um9
Friendly reminder for builders and early-stage founders:
Your friends and family are not your users.
They’re great for support. Terrible for validation.
Talk to people who don’t care whether you succeed, but genuinely care about the problem you’re solving.
Decisions don’t only happen on Zoom.
They happen in rooms, hallways, over coffee, or as a random idea you need to capture before it disappears.
We’re building an iOS companion for @hellonodapp so voice notes can become part of the same team context and knowledge graph as meetings.
Capture it. Share it. Let the team build on it.
Kudos to @dbarabashh for shipping these ideas fast and validating them in product.
WebMCP lets websites expose structured actions directly to AI agents.
Instead of finding buttons and filling forms, an agent can call actions like “search,” “add to cart,” or “create ticket.”
It’s trending because OpenAI just added support to ChatGPT’s built-in browser.
https://t.co/fXnB3jx3rC
AI made every function faster. But teams still work through the same handoffs.
More output doesn’t necessarily mean faster execution. It can mean more coordination overhead.
The next productivity leap will come from redesigning how cross-functional teams work.
Teams and Organizations are coming soon to https://t.co/yp0MQay3Ef.
In the meantime, you can already try Nod for yourself and see how it builds context from your own meetings and conversations.
Most teams don’t have an information problem.
They have a context distribution problem.
That’s a big part of what we’re building with Nod: connect decisions, meetings, work, and blockers so the team doesn’t have to constantly reconstruct context.
Most teams don’t have an information problem.
They have a context distribution problem.
That’s a big part of what we’re building with @hellonodapp: connect decisions, meetings, work, and blockers so the team doesn’t have to constantly reconstruct context.
One of the coolest parts of building Nod is dogfooding what we’re building next.
Shared team context is already changing how we work. Add proactive mode + integrations, and you start seeing what happens when AI doesn’t just remember work, but helps move it forward.
I can’t wait to ship this publicly.