On July 30, we launched SKI (Voice coding for your agents) on Product Hunt, and by the end of the day it became the #1 Product of the Day. It was a moment we had dreamed about for months, and seeing all the hard work finally pay off is something we'll never forget. To everyone who upvoted, left a comment, reported a bug, shared feedback, or simply gave SKI a try, thank you. Every bit of support meant more than you know.
The launch introduced SKI to thousands of people, brought invaluable feedback, improved our visibility across search engines and AI search, and opened doors to opportunities we wouldn't have had otherwise. Beyond the numbers, it reminded us that great products aren't built in isolation. They're shaped by the people who use them.
As we continue building at Pattern AI Labs, we want this community to grow alongside us. That's why we've started a Discord community where we'll share early previews, product updates, behind-the-scenes progress, and new launches. More than that, it's a place for founders, developers, and anyone curious about the future of AI to connect, collaborate, share what they're building, learn from one another, and help shape what we build next.
Talking is fasting than writing. But reading is faster than hearing (at least for me). How does the answering part work? Is it just a status update ("task finished") or does it say the agent's output out loud? We all know that agents can be very verbose in their answers.
SKI
@mateuszkonik Thank you for taking time to look into our product.
You can ask the agent to give a status update or a brief. The default instruction is to limit the talking to 2-3 sentences. But it is customizable and the agent gets to decide it based on user instructions. You could ask it to just share a status update. It will just update on the status on that case. You could ask it to give a brief or open a window and show the issue as well.
It doesn't read out the entire output. This is like a skill which the agent gets to use. It can decide what to tell the user. It can write 100 lines of code, and tell user that it needs your input to work on another task, out loud. So, 100 lines silent, and the request for help is spoken.
SKI
@warmatik ik You’re absolutely right—voice is a great input medium, while visuals are a much better output medium.
SKI
@mateuszkonik We’re are also working on a dynamic UI that adapts to the conversation and presents information visually, which we think will solve this problem.
SKI
@mateuszkonik Thanks for sharing feedback
SKI
@mateuszkonik Thanks for the feedback
Congrats on the launch! The "answers you out loud" part is what got me. most voice tools are one-way. Curious how it handles longer, more ambiguous prompts (not just quick one-liners), does it ask clarifying questions back, or just runs with its best guess?
SKI
@theresa_lindner Thank you for checking out SKI. This is two-way voice, completely local.
Regarding the longer prompts spoken, it depends on the agent. The agent connected to SKI can decide to ask clarifying questions, or to move ahead. SKI works just like the text box in coding agents. The brain is always the connected agent. We have seen that the behaviour depends largely on the way the user using the session.
Some users just vibe code with a end goal given. For them, the agents usually do not ask much clarifying questions. For the people who usually explore deep into the problems and make fine grained decisions, the agents tend to ask before making decisions. But it largely depends on the connected agent, its context and the user using it.
@johnkg003 That makes sense, so SKI is really just the voice layer, the clarifying behavior depends on the user's style and the connected agent. Gonna give this a try, starting with Claude Code I think.
SKI
@theresa_lindner Sure, please let us know your feedback. It will help us improve the product for everyone!
SKI
@theresa_lindner Thanks
SKI
@theresa_lindner Thanks Theresa
SKI
@theresa_lindner 🔥
SKI
@artk Thank you for checking out SKI.
If you put Claude Code in Auto approval mode, you can work full away from screen.
Yes, it is possible to let the agent join a meeting, like Google Meet and talk to it in meeting, or connected to Android auto or Car Play. This is a paid feature. Needs account registration in SKI. Free hours included. Powered by agentcall.dev (which is also our product)
SKI
@artk Thanks for trying SKI
SKI
@artk Thanks again for checking SKI
SKI
@artk 🔥
I run Claude Code all day as a solo dev and the thing that actually breaks my flow is drifting to another tab and not noticing when it's done or sitting blocked on an approval. Voice-out closing that loop is the part I actually want — does it speak up when the agent is waiting on me, or only when a task finishes?
SKI
@lennoxbeflying Thank you for taking a look at SKI.
You can ask the agent to talk as you want it to, and it will use the SKI skill to adapt to your needs. For instance, if you need updates on the stages of completion, it can give that to you. If you want only the completion or waiting-on-you updates, it can do that as well.
SKI functions as a voice capability for your agent to talk to you. The agent you connect to SKI can decide what to speak. You can ask or instruct the agent, and it will act according to your instruction.
Please take a look into SKI. And let us know your feedback or if you have any queries and we would be happy to help :)
SKI
@lennoxbeflying Thanks for trying SKI
SKI
@lennoxbeflying Thanks again for checking SKI
SKI
@lennoxbeflying 🔥
Congrats on the launch. Just a quick question, how does the push-to-talk handle noisy environments or mechanical keyboard chatter during active coding?
SKI
@stacy_liu2 Thanks a lot for checking out SKI. SKI has noise cancellation at a default mode at present. Will be bringing in customizable levels for noise cancellation in the upcoming version.
At present, it is not affected by mechanical keyboard chatter. The issue comes in when the surroundings are too much noisy with multiple people talking out loud.
SKI
@stacy_liu2 Thanks for trying SKI
SKI
@stacy_liu2 Thanks again for checking SKI
SKI
@stacy_liu2 🔥
The teammate framing is fun, having Claude Code answer me out loud feels closer to pairing than prompting. I could see myself using this while pacing around the room thinking through a refactor. Does it hold up in a noisy space like a shared office, or does it really want a quiet room?
SKI
@doganakbulut Thank you for checking out our product. It is like a pair programming. We would recommend wireless headphones in case you are far away from the device. It doesn't necessarily require a quiet room, but a quite room ensures much better transcription accuracy. When you are talking continuously, it usually gets transcription right. The noisy room creates issues only when you are silent and the mic is turned on, where it picks up the low volume talks by someone else. If you are the loudest in the room, it works.
Will try to bring in a noise cancellation slider in the next version where you can adjust the settings depending on a quiet room and a noisy one. Thanks a lot for bringing this up.
SKI
@doganakbulut Thanks for checking
SKI
@doganakbulut 🔥
The part that clicked for me is the agent talking back — most voice tools only fix the input half. I build my products solo after a full-time job, so there's no teammate to think out loud with, and something that answers is a different thing than dictation. One question: can I speak one language and have the prompt come out in another? I'd be talking to it in Spanish while my codebase and prompts are English.
SKI
@soysebalopez Thanks a lot for checking out SKI. In Mac OS, we use Whisper and it supports Spanish. Use transcription settings low to high. The ultra version supports only English for now. If the agent connected to SKI can understand spanish, the setup will work well. It will be as if you typed in Spanish to the agent. We currently do not have a mechanism to convert Spanish to English within the app.
SKI
@soysebalopez Thanks a lot for trying SKI
SKI
@soysebalopez Thanks again for checking SKI
SKI
@soysebalopez 🔥