Does Agent Brain send my code or memories to a cloud service?

No. Agent Brain is self-hosted. Storage, vector search, cache, and local model calls run on your machine or infrastructure.

Which agents can use it?

Claude Code, Codex, Cursor, Gemini, Aider, and custom scripts can connect through MCP or REST.

Memory Governance

Agent Brain is designed for long-lived memory, so governance matters. Without cleanup, any persistent memory system can drift, duplicate itself, or preserve stale decisions.

Governance operations

Agent Brain supports workflows for:

Auditing memories for conflicts and redundancy
Archiving memories that should not appear in default recall
Restoring archived memories when needed
Promoting durable project lessons to broader context
Importing existing agent memory files
Detecting near-duplicates through semantic similarity

Conflict handling

When a new memory is semantically close to an existing one, Agent Brain can flag it for conflict resolution. If the new memory supersedes the old one, the old memory can be archived and linked to the newer decision.

Importing existing memories

Import utilities are available for Claude, Gemini, and Codex memory files. Use them to seed Agent Brain with history before starting a new workflow.

python3 scripts/import-global-memories-claude.py
python3 scripts/import-global-memories-gemini.py
python3 scripts/import-global-memories-codex.py

Project-specific imports should be run from the relevant project root.