Jamie Pine

founder @spacedriveapp - engineer - ui designer - open source ♥️ live daily @

[Jamie Pine @jamiepine]

This is my take on the perfect AI assistant.

A Rust-based agentic operating system designed to scale for large Slack and Discord communities. The channel is the ambassador to the human. Branches think. Workers execute. Nothing ever blocks.

Meet Spacebot 🟣

The biggest issue with OpenClaw is when it's doing work, it can't talk to you. Spacebot's architecture fixes this by design the conversation layer never touches tools. It delegates thinking to branches and heavy tasks to workers, so it's always responsive even with 100 people talking at once.

Dump your memory files, notes, documents and chat histories into a folder — Spacebot turns them into structured memories automatically. Eight typed memory categories, graph associations, hybrid search. Not markdown files. Not vibes in a vector database.

Built-in @OpenCode workers for deep coding sessions. Browser automation. Brave web search. Cron jobs. A skill system compatible with your existing OpenClaw skills. And a gorgeous control UI at https://t.co/aAEbMp4LSh.

The cortex oversees the whole system — auditing memories, actioning goals and todos. You teach your Spacebot by talking to it. Structure and speed over config files and markdown.

Self-hosting is a single Rust binary. Or one-click cloud deploy at https://t.co/aAEbMp4LSh.

This is for teams, communities, and personal assistants. It will blow you away.

⭐️ https://t.co/H8pCoE9R4h

[Jamie Pine @jamiepine]

there is a non zero chance Voicebox was used in the creation of this meme

["Guru Larry" Bundy Jr @LarryBundyJr]

An automated A.I. WWE news channel on YouTube tries to pronounce "WWE" as a word and ends up going insane!🤣

[Jamie Pine @jamiepine]

@KarstenBeoulve Means "software as a service". I usually build open source apps that run on your computer, this is a cloud based service.

[Jamie Pine @jamiepine]

I'm currently developing an old school closed source SaaS, using entirely TypeScript and pnpm. It is so refreshing. You'll hear about that soon, but I want to share how I'm building it.

I've paired the backend with a Spacebot deployment to orchestrate agents per customer, as well as dedicated higher level operations agents. AI employees codified into backend logic using an SDK.

My workflow is Voicebox dictations, because the audio and transcript are saved locally for archive, to create dated research documents, I have 36 of them so far, converted into 142 tasks covering development to MVP and into production scale. The tasks are just in-repo markdown files with a metadata schema and naming convention like "API-001-hono-scaffold.md". They're more like specs, with full implementation guidelines and can evolve with the code. Finally I turn those into official docs when the code is complete and tested.

I watch every coding agent carefully, usually 2-5 at a time and I tackle 2-10 tasks per day depending on the size. I'm working with another backend developer, a close friend, who is cleaning up any slop behind me, every commit. Not using branches, just pushing to main.

I take a little extra time on the planning phase, doing deep research, web scraping. I spend the largest chunks of time on UI, but have a lot of existing material from my other apps like Spacedrive, Voicebox and Spacebot to use as reference. Research is key, you can never do too much. I capture every thought I have as a voice note, and have the agent expand it into a research doc.

Oh and canceled my Claude sub, exclusively using GPT 5.5 through OpenCode, being careful to pick the right thinking complexity for the job, because AI overthinking is a real issue in many tasks. Gippity is killing it right now.

I started this almost three weeks ago, moving exceptionally fast I was even able to demo it at Web Summit Vancouver, but in the last four days alone I've built the majority of the platform. I turned the docs into a deck and plugged the data room as an agent chat along side the deck, so potential investors can ask questions before chatting with me.

Feels like the golden age to build a new platform. Pretty hyped about this one.

[Umbrel ☂️ @umbrel]

sure, an ai agent running on your home server is cool but how about a team of them running in parallel, sharing memory, handling your discord, your team slack, your community, all from the umbrel on your shelf?

@spacedriveapp's spacebot is now live on the umbrel app store, lfggg

[Sammy Lau | Personal Finance @SammyLau]

@araghougassian Vancouver on a summer day is one of the best places on Earth.

[Jamie Pine @jamiepine]

@arugulapiquante @604RAW ahahaha perfect

[Jamie Pine @jamiepine]

I’m not dead, I’m brewing.

If you’re in Vancouver next week for Web Summit hmu (also if anyone can help me get tickets that’d be awesome)

Otherwise catch me at some side events and loitering around the city with Claude

[Jamie Pine @jamiepine]

@icanvardar you got me https://t.co/VSzjWinfez

[Jamie Pine @jamiepine]

@theCTO and that’s why we use it

[thehype. @thehypedotnews]

what’s trending on github this week: hermes agent and claude-mem retain their positions in the weekly ranking for the third time in a row, with 22.8% and 15.8% growth respectively

here are the highest-starred github repos of the week:

andrej-karpathy-skills (@karpathy) +40,732
hermes agent (@NousResearch) +25,081
claude-mem (@Claude_Memory) +10,356
voicebox (@jamiepine) +5,198
evolver (@EvoMapAI) +4,376
genericagent +4,223

agentic ai and persistent memory aren’t slowing down - yet the picture seems far more versatile than the last week.excited to find out where this goes!

[Jamie Pine @jamiepine]

@Old_Vile_Inn restart fully, it shouldn’t be as it works for me again.

if so, i need to know which path is triggering the error

[Jamie Pine @jamiepine]

Voicebox is approaching 20k stars, currently trending on GitHub, has 334k downloads and just shipped 0.4 now supporting 10 models.

On its way to becoming, if not already, the defacto desktop app for voice cloning.

Very proud of it.

[Star History @StarHistoryHQ]

18.8K⭐ voicebox - Open-source voice synthesis studio — clone, generate, and fine-tune voices locally with a clean UI and no API keys required 🎙by @jamiepine

https://t.co/bTxM7kD4CG

#starhistory #GitHub #OpenSource #AI

[Jamie Pine @jamiepine]

@vince_parulan @AdolfoUsier as soon as those models come out, i will add them to Voicebox

[Jamie Pine @jamiepine]

@so_ainsight ありがとう!!

[Jamie Pine @jamiepine]

@sflorimm quite a dumb question, don’t you think?

[Jamie Pine @jamiepine]

@jaezun_ @ErickSky contemplating if i should make an iOS version. not sure the models would work on a phone so it would need to connect to a desktop version, but then, what’s the use case?

[Jamie Pine @jamiepine]

@ProEvilz @ErickSky helpful if you share the error rather than just saying “doesn’t work”. it works for thousands of users.

[Erick @ErickSky]

Adiós ElevenLabs… ya llegó su reemplazo GRATIS Y LOCAL

[VoiceBox]

Con solo unos segundos de audio puedes:

- Clonar cualquier voz en segundos
- 23 idiomas
- 5 motores TTS + efectos de audio
- Timeline tipo DAW para podcasts y conversaciones completas
- 100% en tu máquina (nada sale de tu PC)

Creadores, podcasters, YouTubers y devs de IA lo van a adoptar masivamente.

REPOOO👇

[Jamie Pine @jamiepine]

@KarstenBeoulve It’s only as good as the models it supports, I add more models, you get more options. Voicebox itself doesn’t do anything but provide you a nice interface to use existing models on the market.

[Jamie Pine @jamiepine]

@ApplyWiseAi a) it’s a desktop app, meaning normies can use it easily
b) simple, elegant interface
c) built in DAW editor
d) comes up first when you google “voicebox”, took me a hot minute to find tortoise just now