Brethof-Brain

Brethof-Brain is an AI developer tool that gives AI agents persistent memory across sessions, so an assistant can "remember" prior conversations, decisions, and project context instead of re-explaining everything each time. It is built for developers and teams running AI agents who want long-term context without handing user data to a third-party server. The memory is organized into distinct layers rather than one flat log: curated records that surface current project facts alongside relevant messages, standing rules that load with most prompts, a full searchable chat history, and a decision graph that stores dated decisions with their reasoning and the alternatives that were considered. Notes and playbooks capture agent handovers and reusable procedures, and a background "Ledger" tracks important details automatically. Multiple agents can share one memory store. Privacy is the main design axis: you can run it locally on your own machine (via Docker or Podman, needing roughly 3GB of disk) or use the hosted option, where data is encrypted in Germany under a passphrase you control and processing runs in a secure compute instance in Zurich with zero model-provider data retention. For teams on BestAIFor comparing agent-memory infrastructure, Brethof-Brain stands out for keeping recall local or self-controlled rather than vendor-held.

Giving AI agents long-term, privacy-first memory of conversations, decisions, and project context across sessions.