Sara ghcr.io/farena-ai/sara:latest https://github.com/farena-ai/sara/pkgs/container/sara bridge bash false https://github.com/farena-ai/sara/issues https://sara.backstagecoding.com/ https://github.com/farena-ai/sara/blob/main/docs/INSTALL.md Sara is a self-hosted AI companion and coding partner: chat, memory, a planner, coding tools, photos and voice, all on your own server. You choose the AI model (local with Ollama, or your own provider account); your data and keys stay in her data folder. First start: open the WebUI within an hour and set Sara up in the browser with the 8-character setup code from the container's log (click the icon, then Logs; look for "setup code"). She then shows your owner token once: keep it in your password manager. Missed the hour? Restart the container for a new code. Lost the token later? docker exec Sara node src/main.ts --reset-token (then restart the container). Plain HTTP is fine on your home network; for access from outside, put an HTTPS reverse proxy in front and set Secure cookies and Trust proxy to 1. Productivity: Tools:Utilities http://[IP]:[PORT:7860]/ https://raw.githubusercontent.com/farena-ai/unraid-templates/main/templates/sara.xml https://raw.githubusercontent.com/farena-ai/sara/main/web/icon-512.png --cap-drop=ALL --cap-add=SETUID --cap-add=SETGID --cap-add=CHOWN --cap-add=FOWNER --cap-add=DAC_OVERRIDE --cap-add=KILL --security-opt=no-new-privileges:true --log-opt max-size=10m --log-opt max-file=3 MIT ai assistant companion chatbot llm ollama coding self-hosted https://raw.githubusercontent.com/farena-ai/unraid-templates/main/screenshots/sara-1.png https://raw.githubusercontent.com/farena-ai/unraid-templates/main/screenshots/sara-2.png https://raw.githubusercontent.com/farena-ai/unraid-templates/main/screenshots/sara-3.png An AI model: Ollama (another container or machine) or an account with an AI provider. About 300 MB of memory for Sara herself. 7860 /mnt/user/appdata/sara 0 0