An open-source AI assistant called Clawdbot has quietly amassed over 8,000 GitHub stars and earned coverage from MacStories, with multiple Medium posts going viral this weekend.
Key Takeaways
- Steinberger’s open-source Clawdbot pulled 8,000+ stars as a self-hosted Jarvis.
- It connects to Telegram, WhatsApp, iMessage, and Slack as a single contact.
- Persistent memory is the big draw vs. forgetful consumer AI assistants.
Created by Peter Steinberger, founder of the iOS development company PSPDFKit, Clawdbot runs locally on your computer while connecting to messaging platforms like Telegram, WhatsApp, iMessage and Slack. Users chat with it like a contact in their existing apps.
“To say that Clawdbot has fundamentally altered my perspective of what it means to have an intelligent, personal AI assistant in 2026 would be an understatement,” wrote Federico Viticci at MacStories.
The project solves a persistent problem with consumer AI tools: they forget everything between sessions. Clawdbot maintains memory, preferences and context in local Markdown files that persist indefinitely.
More importantly for power users, Clawdbot can execute shell commands, write and run scripts, control smart home devices and install new capabilities on the fly. Viticci reported burning through 180 million tokens experimenting with it.
For newsrooms, the implications are worth watching. An AI assistant that remembers your beats, sources and research workflows — and runs on your own infrastructure — addresses both the productivity promise and the data privacy concerns that have made enterprise AI adoption complicated. The 2026 Reuters Institute predictions forecast exactly this kind of agentic AI becoming central to newsroom operations.
The catch: Clawdbot requires technical setup and your own API keys from providers like Anthropic or OpenAI. It’s a tinkerer’s tool, not a consumer product. But its rapid growth suggests demand for AI assistants that users actually control.
“2026 is already the year of personal agents,” one user wrote on the project’s website.
Clawdbot is available free on GitHub. Documentation lives at docs.clawd.bot.
