Skip to content
 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

10 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Motion Memory Core

Persistent memory for AI agents — two tiers, one system.

Your AI has amnesia. Every new session starts from zero: it doesn't know you corrected it last week, doesn't know you're under a merge freeze, doesn't know you hate mocks in tests. Motion Memory Core fixes that with two complementary tiers you can adopt independently or together.

Local Memory MCP Memory Server
Answers "What should this agent know in this repo?" "What should I remember, everywhere?"
Storage Markdown files, committed to the repo Postgres, self-hosted, MIT
Scope One repository Every client, every machine
Setup Copy templates, paste a prompt — no infra docker compose up, one bearer token
Works with Claude Code, Cursor, any file-reading agent BoltAI, Claude Desktop, Cursor, Claude Code, any MCP client — over native Streamable HTTP

Read docs/architecture.md for how the two tiers relate and when to use each — short version: Local Memory for what's specific to a codebase and worth a PR review; the MCP server for durable facts about you that should follow you into every repo and every tool.

Quickstart

Local Memory — no infrastructure, start in a repo right now:

cp -r local/templates .claude/projects/$(pwd | sed 's|/|-|g')/memory
# then paste local/prompts/CLAUDE.md-snippet.md into your CLAUDE.md

Full guide: local/README.md.

MCP Memory Server — self-hosted, one memory for every AI client:

cd server && cp .env.example .env   # set STASH_TOKEN + provider key
docker compose up -d

Full guide: server/README.md · Connect any client: server/CONNECTING.md · One-click deploy: server/deploy/railway.md.

The 4 memory types

Both tiers share the same typed model — user, feedback, project, reference — so the mental model is identical whether a memory lives in a markdown file or a database row. Reference: docs/types-reference.md · docs/when-to-save.md.

License & credits

  • Motion Memory Core (this repo): MIT © 2026 Tolowa Studio (see LICENSE).
  • The MCP server's memory engine is Stash (Apache‑2.0) — https://github.com/alash3al/stash — used as a dependency, not vendored. See NOTICE. We've contributed the native Streamable HTTP transport upstream: alash3al/stash#15.

About

An open, self-hostable memory layer for any AI agent. One memory, every client. MIT. Powered by the Stash engine.

Resources

Stars

Watchers

Forks

Releases

Packages

Contributors

Languages