Clawdbot
lastmod 2026-02-04created 2026-01-264 edits
A personal AI assistant that works on every OS and platform
About
- An LLM-based agent that runs on the user's computer, with shell & file system access and the ability to control the computer
- Can control everything: inbox cleanup, sending mail, calendar management, browsing (input & data extraction), etc.
- Chat through most messaging platforms you already use, without installing a separate app (Slack, Telegram, Discord, WhatsApp, ..)
- All settings, preferences, user memory, and instructions are stored as folders and Markdown documents, so they are freely customizable (you can even ask Clawdbot to change the settings)
- Runs on macOS, Linux, Windows (WSL2)
- Supports integration with Anthropic Claude Pro/Max and OpenAI ChatGPT/Codex models
- Any model works, including Gemini/xAI/OpenRouter/Mistral/DeepSeek/GLM and Ollama/LM Studio
- WhatsApp, Telegram, Slack, Discord, Google Chat, Signal, iMessage, Microsoft Teams, WebChat
- Productivity: Apple Notes, Apple Reminders, Things3, Notion, Obsidian, Bear Notes, Trello, GitHub
- Music : Spotify, Sonos, Shazam
- Smart home : Philips Hue, 8Sleep, Home Assistant
- Tools and automation: Browser (Chrome/Chromium), Canvas (A2UI), Voice (VoiceWake + Talkmode), Gmail, Cron, Webhooks, 1Password, Weather
- Media : Image Gen, GIF Search, Peekaboo (screen capture), Camera (photo/video capture)
- Social : X (tweet, reply, search), Email (send and read email )
See also
Favorite site
- Clawdbot — Personal AI Assistant
- Clawdbot - A personal AI assistant that works on every OS and platform | GeekNews
Articles
- How Clawdbot manages memory: a Markdown-based hybrid memory architecture | GeekNews
- [Original] How Clawdbot Remembers Everything - Manthan
- Clawdbot is an open-source local AI assistant that stores and manages long-term memory as Markdown files in the user's local environment, with no cloud dependency.
- A two-layer memory system (daily logs and long-term knowledge) and hybrid search (vector + keyword) enable efficient information retrieval even over large contexts.
- To overcome context window limits, it uses 'Compaction' and 'Pruning' techniques, achieving cost optimization while maintaining performance.
Pages that link here