Memory across machines
A self-hosted hub connects your peers to synchronize working context. You choose the environment and configure access.
A PROJECT BY GUILHERME PINHEIRO · DATANO
Animus
Animus collects, indexes and synchronizes AI assistant memories across your machines. Retrieve conversations, find earlier decisions and provide context to agents through the Model Context Protocol.
Open source · GNU AGPLv3
Source available on GitHub, created by Guilherme Pinheiro / Datano and licensed under GNU AGPLv3.
A self-hosted hub connects your peers to synchronize working context. You choose the environment and configure access.
Find conversations and decisions using SQLite FTS5 full-text search and context retrieval capabilities.
Expose query tools to your assistant within the context and permissions of your installation.
The experimental integration supports opt-in curation. Exported notes start as candidates; verified knowledge and training have their own requirements.
Opening the code does not expose user memories. Install your own hub, configure credentials and HTTPS, and select only sources you are authorized to collect and synchronize. Automatic filtering does not replace human review of personal or corporate information.
Created and maintained by Guilherme Pinheiro in the Datano ecosystem. The current code uses AGPL-3.0-only: retain copyright and license notices, identify modifications and meet the source availability obligations, including for modified versions accessed over a network.
Forks and commercial use are permitted under the license. Previously distributed MIT versions retain their permissions. The code grants no rights to third-party trademarks and implies no endorsement by the official distribution.
Read the licenseAn independent project, unaffiliated with Ubisoft, assistant vendors or Obsidian.
Loading security verification…
Protected by Cloudflare Turnstile. Privacy information.