🚀🔥 Big update Feb 2026: Channels + DMs are now live on Starships.ai

This is a major step forward from what we had before.

You can now organize conversations into channels inside each Starship, and use DMs for quick 1:1 chats 💬 No more messy threads. No more lost context. Everything stays in one place.

Along with this, we’ve shipped a lot more in this release:

✨ Reactions + emojis 🔔 Better notifications (with sound alerts + a dedicated page) 🟢 Online status indicators 📱 Installable app (PWA support) 🧭 New onboarding with smarter default setup 🛠️ Marketplace and deploy upgrades ⚡ Major stability and performance improvements

Our focus is simple: Help you move faster, collaborate better, and turn ideas into real products without friction 🔁

This update lays the foundation for how teams will work inside Starships going forward.

https://starships.ai/ Try it out, push it, and tell us what breaks. Your feedback directly shapes what we build next 🙌

Welcome to Starships.ai 🚀

I’m Rudra.

I spend most of my time building, but a surprising amount of it still goes into running things: posting updates, tracking what worked, checking SEO, following up on loose ends, and stitching tools together so work actually gets done.

AI helped with pieces of this, but never the whole thing.

The problems I kept running into

• “Someone still has to run it” work
  AI can draft, but scheduling, tracking results, deciding next steps, and closing the loop still fall on me.

• Everything lives in a different tab
  Content, analytics, decisions, and execution stay disconnected, so workflows never actually run end to end.

• Nothing improves with reuse
  The same growth, SEO, and ops tasks repeat every week without memory or compounding.

I ran into these while trying to automate my own growth, ops, and product work.

So I started building something I actually needed.

What we’re building

Starships.ai is an internal-first system for automating end-to-end work with AI agents.

Instead of single prompts or isolated tools, Starships.ai lets us:

• Pass work between agents such as research to content to publishing
• Run workflows with real steps like fetch data, decide, act, retry, and complete
• Let agents use real tools including APIs, browsers, code, and internal systems
• See exactly what happened in every run, including failures and retries

If it cannot run reliably without constant human intervention, it does not count as automated.

What we want to achieve

The goal is simple.
Let agents own repetitive work so humans stay focused on direction and judgment.

That means:

• Less context switching
• Fewer manual loops
• Stepping in for oversight, not execution

If it does not save time every week, it is not done.

What we’ll share here

This channel is where I’ll share how I’m using Starships.ai internally:

• End-to-end workflows we automate
• What worked, what broke, and why
• Tradeoffs made while building reliability
• Changes shipped after real usage

Updates will be concrete, example-driven, and based on things I actually run.

If this sounds familiar and a lot of your work is still manual, feel free to drop a message. Happy to help simplify it.

Link to the project https://starships.ai