We're excited to announce v0.10.21 with several major features and improvements.
Voice to Text (Preview) Transcribe voice input in the Agent sidebar entirely on-device. Your voice data never leaves your machine, giving you complete privacy and control over your transcriptions.
One week since launch, and we're incredibly grateful for all the support. Thank you to everyone who upvoted, tried the tool, and shared feedback - it means the world to us. We've been heads-down shipping since day one. Here's what landed in just 7 days:
Multi-platform publishing - Publish directly to Instagram, Threads, and X - not just LinkedIn anymore
Guided Mode - The AI walks you through the right questions before generating, so your carousels are better from the start
Built-in web research - Paste any URL (blog post, article, landing page) and the AI reads it to create your carousel
Better generation & editing quality - Overall improvements to how carousels are generated and edited
Bug fixes - We listened and fixed several issues reported by our users
We're building this for you, and we mean it. Your feedback drives every decision we make.
Most software wants you to come back every day. The business model depends on it. More sessions, more engagement, more opportunities to monetize.
But what happens when your product's purpose is to help someone understand themselves better? At Murror, we've been wrestling with a paradox: if we do our job well, users should eventually need us less not more.
ClawOffice was a bit of a gimmick - a 3D virtual office for AI agents. It was fun to build and got some attention on launch, but let's be real: it didn't take off. People thought it was cool for a minute, then moved on. No real retention, no real problem being solved.
Here's what I actually learned from it:
Novelty value. A cool concept gets you a launch day. It doesn't get you users who come back on day 30.
I was building for the demo, not the workflow. ClawOffice looked great in a screenshot. It didn't solve anything measurable for anyone.
"What gets tracked gets improved" is real. The founders I talked to afterward all had the same pain - they were shipping features and running experiments with no clue what was actually driving revenue.
Sleep debt doesn't just make you tired. It degrades decision-making, reaction time, and emotional regulation, and you often can't feel how impaired you are.
Most people try to fix it with more caffeine. That makes it worse.
Clarity helps you use caffeine to perform better today without destroying tomorrow.
5 days until our Product Hunt re-launch. Follow along. We're excited to show you the changes since the last launch.
We just went live with the SigmaMind MCP Server, and we re on a mission to end "infrastructure hell" for voice developers.
For the last year at SigmaMind (YC S22), we ve watched builders struggle to stitch together telephony, low-latency models, and fragmented APIs. Today, we re changing that. We ve built a way to configure and deploy production-grade voice agents directly from your IDE (Cursor, Claude Code, etc.) using the Model Context Protocol. No more manual glue - just one prompt to connect your model, pick a voice, and get a live phone number.
Hey everyone - quick update on DeployHermes (managed hosting for Hermes agents on Fly.io).
Since our last public release (right after we moved the stack to Vercel), we ve been heads-down on reliability and on features people actually asked for. Here s what s new:
Lately, I ve been looking closely at how independent builders and small teams are managing AI knowledge bases. It feels like the default "industry standard" is to immediately reach for a complex RAG pipeline and a heavy, paid Vector Database.
But I'm starting to wonder if we are over-engineering this for 90% of standard use cases.
Vector DBs are incredibly powerful for massive scale, but for smaller or non-massive datasets, they can be expensive, complex to query, and act as complete black boxes. If a search returns a weird chunk, diagnosing it is often a nightmare.
Every AI product I see launching right now is racing to add the most impressive, most complex AI feature they can build. Autonomous agents. Multi-step reasoning. Real-time analysis of everything.
When we started building Murror, we fell into the same trap. We wanted to build the smartest emotional AI possible. Something that could analyze patterns across months of conversations, predict emotional states, generate deep psychological insights.
Every time I built a dashboard for one client, I had to rebuild it for the next. Creating clients, reports, dashboards, data sources felt like clicking through ten different pages just to do one simple thing
So we made some changes
Now in ZapDigits
You can create everything from the sidebar. No more jumping around Any dashboard can become a template you reuse Templates have their own gallery so you can see yours and ours in one place More Google Analytics metrics to get better insights Each dashboard can have its own look
Everyone tells you to ship fast. Move fast and break things. Get to market before someone else does.
I believed this for a long time. When we were building Murror, speed was everything. We pushed features weekly, sometimes daily. We celebrated every deploy like a small victory.
