Let's have a sticky for Clawdbot talk. We need to really dig into this.
https://clawd.bot/
https://github.com/clawdbot/clawdbot
AI Quote: Clawdbot is an open-source, self-hosted personal AI assistant that runs entirely on your own hardware—whether that's a Mac, Linux machine, Windows via WSL, Raspberry Pi, or even a dedicated VPS. Launched in early January 2026 by developer Peter Steinberger (founder of PSPDFKit/Nutrient), it has rapidly gone viral in the AI community, amassing thousands of GitHub stars and a dedicated Discord. Unlike cloud-based chatbots from big tech, Clawdbot keeps all your data, conversations, and context private and local, with no uploads to external servers. It integrates seamlessly into the messaging apps you already use—like WhatsApp, Telegram, Discord, Slack, Signal, or iMessage—turning it into a contact you can text naturally, making it feel like chatting with a super-capable teammate or "AI employee" rather than switching to yet another app.
What sets Clawdbot apart is its ability to actually perform real-world actions instead of just responding with text. It can clear your inbox by unsubscribing from spam, send emails on your behalf, manage your calendar (rescheduling conflicts autonomously), check you in for flights, browse the web to fill forms or extract information, run shell commands and scripts, control smart home devices, handle code tasks like debugging or opening GitHub PRs, fetch health stats from apps like WHOOP, generate summaries from meetings or notes, and even spawn background sub-agents for research. It supports persistent memory across sessions (often stored in local Markdown files), learns your preferences over time, and can be proactive—sending morning briefings with weather, objectives, reminders, trending topics, or personalized content suggestions without you asking first.
Powered by leading LLMs like Claude (Anthropic), GPT models, or fully local ones via Ollama (for complete offline use), Clawdbot is highly extensible through community-built "skills" and plugins, allowing it to connect to services like Gmail, GitHub, Todoist, Spotify, Obsidian, and more. Users describe it as the closest thing yet to a true "Jarvis"-style personal AI OS or butler—hackable, autonomous, and transformative for productivity, with many running it 24/7 on dedicated hardware like a Mac Mini. While it's still early and can require careful setup for security (e.g., sandboxing tools, avoiding full access to sensitive accounts), its open-source nature (MIT license) and rapid community momentum make it a standout in the shift toward owning and controlling your own AI agents. 🦞
Who has one of these setup? I am trying to decide if I want to build a PC from old parts, use an old server I have laying around, or buy something cheap for $200.
This guy has a nice thread about cheap hardware options.
https://clawd.bot/
https://github.com/clawdbot/clawdbot
AI Quote: Clawdbot is an open-source, self-hosted personal AI assistant that runs entirely on your own hardware—whether that's a Mac, Linux machine, Windows via WSL, Raspberry Pi, or even a dedicated VPS. Launched in early January 2026 by developer Peter Steinberger (founder of PSPDFKit/Nutrient), it has rapidly gone viral in the AI community, amassing thousands of GitHub stars and a dedicated Discord. Unlike cloud-based chatbots from big tech, Clawdbot keeps all your data, conversations, and context private and local, with no uploads to external servers. It integrates seamlessly into the messaging apps you already use—like WhatsApp, Telegram, Discord, Slack, Signal, or iMessage—turning it into a contact you can text naturally, making it feel like chatting with a super-capable teammate or "AI employee" rather than switching to yet another app.
What sets Clawdbot apart is its ability to actually perform real-world actions instead of just responding with text. It can clear your inbox by unsubscribing from spam, send emails on your behalf, manage your calendar (rescheduling conflicts autonomously), check you in for flights, browse the web to fill forms or extract information, run shell commands and scripts, control smart home devices, handle code tasks like debugging or opening GitHub PRs, fetch health stats from apps like WHOOP, generate summaries from meetings or notes, and even spawn background sub-agents for research. It supports persistent memory across sessions (often stored in local Markdown files), learns your preferences over time, and can be proactive—sending morning briefings with weather, objectives, reminders, trending topics, or personalized content suggestions without you asking first.
Powered by leading LLMs like Claude (Anthropic), GPT models, or fully local ones via Ollama (for complete offline use), Clawdbot is highly extensible through community-built "skills" and plugins, allowing it to connect to services like Gmail, GitHub, Todoist, Spotify, Obsidian, and more. Users describe it as the closest thing yet to a true "Jarvis"-style personal AI OS or butler—hackable, autonomous, and transformative for productivity, with many running it 24/7 on dedicated hardware like a Mac Mini. While it's still early and can require careful setup for security (e.g., sandboxing tools, avoiding full access to sensitive accounts), its open-source nature (MIT license) and rapid community momentum make it a standout in the shift toward owning and controlling your own AI agents. 🦞
Who has one of these setup? I am trying to decide if I want to build a PC from old parts, use an old server I have laying around, or buy something cheap for $200.
This guy has a nice thread about cheap hardware options.



![[Image: vibesignature.gif]](https://i.ibb.co/8LvxwtYQ/vibesignature.gif)


![[Image: 6KMyuYs.png]](https://i.imgur.com/6KMyuYs.png)


