memgit 0.2.0

Git for AI memory — version-controlled context persistence across Claude, GPT, Gemini, Cursor, Windsurf, and more

memgit — git for AI memory

Your AI assistants forget everything when the session ends. memgit fixes that.

Version-controlled, cross-AI context that persists, diffs, rolls back, and syncs like code. Switch from Claude to Cursor to ChatGPT mid-project — your context is already there.

---

Why not claude.md? Why not mem-search?

You've probably already tried both. Here's why they hit a ceiling:

Capability	claude.md	mem-search plugin	memgit
Loads only relevant context	❌ loads everything	⚠️ loads recent observations	✅ BM25 search — top-k per query
Version history	❌	❌	✅ full commit log
Diff between sessions	❌	❌	✅ memgit diff
Roll back a wrong memory	❌ manual edit	❌	✅ memgit rollback
Works in Cursor, Windsurf, GPT	❌ Claude only	❌ Claude only	✅ all via MCP / HTTP
Team sync	❌ copy-paste files	❌	✅ memgit git push
Scales to 10k+ sessions	❌ file grows	❌ search slows	✅ memgit squash
Measurable token savings	❌	❌	✅ memgit stats
Export / import standard format	❌	❌	✅ TOON + git

---

Proof — token savings you can measure

Run this on your own store to see the actual numbers:

$ memgit stats

  Total memories:   108   (41 feedback · 23 user · 19 project · 12 reference · 8 convention · 5 lesson)
  Priority:          3 critical · 67 medium · 38 low

  Token cost comparison:
  ┌─────────────────────────────────────┬──────────────────┬───────────────────┬─────────────────────┐
  │ Approach                            │ Tokens/session   │ vs full load      │ $/session (GPT-4o)  │
  ├─────────────────────────────────────┼──────────────────┼───────────────────┼─────────────────────┤
  │ claude.md / dump all memories       │ 12,840           │ 100%  baseline    │ $0.0321             │
  │ memgit search (BM25 top-8)          │ 640              │ 5%  (95% savings) │ $0.0016             │
  └─────────────────────────────────────┴──────────────────┴───────────────────┴─────────────────────┘

  Weekly savings (10 sessions/week):
    Tokens saved:   122,000/week
    Cost saved:     $0.31/week  →  $15.86/year  (at GPT-4o input pricing, $2.50/M)

Why such a big difference? claude.md loads all context every session. memgit uses BM25 relevance scoring — it loads only the 8 memories most relevant to the current session, not everything you've ever recorded.

---

The git analogy is literal

memgit's data model maps exactly to git:

memgit	git
mnemonic	file
MindState	tree
checkpoint	commit
thread	branch
memgit commit	git commit
memgit diff	git diff
memgit log	git log
memgit squash --keep-last 100	git rebase -i --autosquash
memgit git push	git push

This is not metaphorical — memgit uses a content-addressed object store (SHA-256 blobs) identical to git's architecture. Every memory has a stable SHA. Identical content has identical SHAs. Old state is always recoverable.

---

The store IS a git repo

Every memory is a readable .toon file under memories/. Push your entire memory set to GitHub with standard git:

memgit git init --remote [e-mail]:yourteam/ai-memory.git
memgit git push

Teammates pull and start with your AI's learned rules from session 1:

git clone [e-mail]:yourteam/ai-memory.git ~/.claude/memgit-store
memgit setup all

You can grep, git blame, and git diff your memories just like code:

grep -rl "database" ~/.claude/memgit-store/memories/
git log --follow memories/no-db-mock.toon
git diff HEAD~7 memories/

---

Install

Mac / Linux:

pip install memgit

Mac (Homebrew):

brew tap code4161/tap && brew install memgit

Windows:

pip install memgit

(choco install memgit is not live yet — the Chocolatey package is not on community.chocolatey.org. Use pip until it lands.)

Any AI tool config (no Python needed — npx auto-installs on first run):

{ "mcpServers": { "memgit": { "command": "npx", "args": ["-y", "memgit-mcp"] } } }

---

Quickstart (3 minutes)

# 1. Install and initialize
pip install memgit
memgit init               # auto-detects best location (~/.claude/memgit-store etc.)

# 2. Import existing memories (if you use Claude Code)
memgit import claude-code ~/.claude/projects/

# 3. Register with your AI tools (interactive picker)
memgit setup

# 4. See your token savings
memgit stats

Restart your AI tool — it now searches your memory store at the start of every session.

---

Resume where you left off

Ask an AI "can we proceed on the pending tasks?" in a fresh session and it will guess from whatever file happens to be open. memgit resume replaces the guess with the record:

memgit resume            # last checkpoints, work in flight, recent + critical memories
memgit resume --plain    # plain text, for piping into an AI context
memgit resume --json     # for tooling

Wire it into Claude Code so every new session starts with this digest in context — no tool call, no judgment required:

memgit setup hooks       # installs a SessionStart hook (~/.claude/settings.json)

The digest is deliberately bounded (~350 tokens measured on a 500-memory store): rules are clipped, the critical list is capped, and full text is one get_memory call away.

---

Scale to 10,000+ sessions

After months of use, your checkpoint history grows. Squash compresses it, gc reclaims the disk:

memgit squash --keep-last 100    # keep last 100 checkpoints, squash everything older
memgit squash --older-than 30    # squash everything older than 30 days
memgit squash --dry-run          # preview first

memgit gc                        # delete unreachable objects, trim reflogs
memgit gc --dry-run              # preview
memgit gc --squash-keep 200      # compact history, then sweep

The current memory state is always preserved — and squash is lossless-in-substance: every collapsed checkpoint leaves a one-line record (time, author, diff, message) in an append-only archive under .memgit/logs/archive/ that gc never touches. Benchmark on a 2,000-checkpoint store: 94% smaller (39.5 MB → 2.2 MB), fsck clean. History operations stay O(1) as the chain grows (SHA resolution and checkpoint counting measured at ~0.08 ms at 2,000 checkpoints).

---

Multiple agents, one memory

All writes go through a git-style store lock (0.08 ms overhead), so concurrent agents can't corrupt the store or lose each other's updates. Two patterns:

Shared thread — agents write concurrently; if one commits while another has work staged, the second commit auto-merges (three-way, against the recorded base) instead of clobbering. Set MEMGIT_AUTHOR=agent-name so each checkpoint says who did it.

Thread per agent — isolate, then integrate:

memgit thread create agent-1     # branch off for each agent
# ... agents work on their own threads ...
memgit merge agent-1             # three-way merge back (common-ancestor based)

Conflicts (same memory changed on both sides) resolve to the newest version; an edit always beats a delete. Both histories are preserved.

---

What the AI sees

Once registered via MCP, every AI tool gets 6 tools:

Tool	When the AI uses it
resume_session	When the request depends on prior state — "continue", "the pending tasks", session start
search_memories	Before answering anything that touches past work or preferences
get_memory	When it needs full details of a specific memory
list_memories	To browse or audit what's stored
save_memory	When it learns something worth keeping for next time
get_checkpoint_log	To check when memories were last synced

The tool descriptions teach the AI judgment — "does this request depend on state you don't have in context?" — rather than keyword triggers. Measured cost of the whole tool surface: ~1,150 tokens once per session; a resume_session reply is ~335.

---

Commands

# Core (git-like)
memgit init                       # initialize store (auto-detects best path)
memgit add <slug> <rule>          # stage a memory
memgit commit -m "message"        # checkpoint current state
memgit log                        # history
memgit diff [sha1] [sha2]         # what changed
memgit show <slug>                # display a memory
memgit remove <slug>              # remove from active index (history preserved)
memgit status                     # staged changes
memgit search <query>             # BM25 relevance search
memgit rollback <ref>             # restore state to a checkpoint (HEAD~N or SHA)
memgit resume                     # where we left off — session-start digest
memgit merge <thread>             # three-way merge a thread into the current one

# Scale & proof
memgit squash                     # compress old history (archives what it collapses)
memgit gc                         # reclaim disk: sweep unreachable objects
memgit stats                      # token savings + disk usage
memgit lint                       # validate all memories
memgit fsck                       # verify store integrity

# Import / export
memgit sync                       # sync from Claude Code files + commit
memgit import claude-code <path>
memgit import file <path>
memgit export <slug>

# Git sync (team features)
memgit git init [--remote URL]
memgit git push [remote] [branch]
memgit git pull [remote] [branch]
memgit git export
memgit git status

# AI tool registration
memgit setup                      # interactive step-by-step picker
memgit setup all                  # auto-register every detected tool
memgit setup claude-code
memgit setup cursor
memgit setup windsurf
memgit setup cline
memgit setup continue
memgit setup gemini-cli
memgit setup hooks                # Claude Code SessionStart hook → auto-inject resume digest

# Server
memgit serve                      # MCP stdio (Claude Code, Cursor, Windsurf, Cline)
memgit serve --http               # HTTP REST (ChatGPT Custom Actions, Gemini)

# Visualization
memgit graph                      # D3.js interactive relationship map
memgit thread list / switch / create

---

AI tool support

Tool	Protocol	Command
Claude Code	MCP stdio	memgit setup claude-code
Claude Desktop	MCP stdio	memgit setup claude-desktop
Cursor	MCP stdio	memgit setup cursor
Windsurf	MCP stdio	memgit setup windsurf
Cline / Roo-Code	MCP stdio	memgit setup cline
Continue.dev	MCP stdio	memgit setup continue
ChatGPT (Custom Actions)	HTTP + OpenAPI	memgit serve --http → import http://localhost:7474/openapi.json
Gemini API	HTTP function calling	memgit serve --http + llm-tool-definitions.json
Any MCP tool	MCP stdio	Add {"command": "memgit", "args": ["serve"]} to config

---

TOON format — compact, readable, diffable

Standard markdown memory file:

## Rule: Never mock the database in tests
**Type:** feedback  
**Priority:** medium  
**Why:** We got burned last quarter — mocked tests passed but the prod migration failed.  
**When to apply:** Any time writing tests that touch persistence layers.  
**Tags:** testing, database

The same memory in TOON:

TOON1|fb|no-db-mock|2026-07-01T10:00Z
#testing #database
RULE:Never mock the database in tests
WHY:Mocked tests passed but prod migration failed last quarter
WHEN:Any persistence test

Measured with a real tokenizer, TOON is ~5–10% leaner than equivalent markdown — a nice bonus, not the headline. The headline saving is retrieval: memgit loads the top-8 relevant memories per query instead of everything.

At 108 memories: 12,840 tokens (dump everything) → 640 tokens (memgit BM25 top-8)

For exact token counts in memgit stats, install the optional tokenizer: pip install "memgit[tokens]".

---

Architecture

~/.claude/memgit-store/
  .memgit/
    objects/     ← SHA-256 content-addressed blobs (gzip compressed)
    refs/threads/main   ← HEAD checkpoint SHA
    TOON_INDEX   ← active slug→sha mapping
    config       ← author, default thread
    logs/        ← ref change audit trail
  memories/      ← flat .toon files (git-trackable, human-readable)
  .git/          ← standard git repo (after `memgit git init`)

---

Contributing

git clone https://github.com/code4161/memgit.git
cd memgit
python -m venv .venv && source .venv/bin/activate
pip install -e ".[dev]"
pytest    # 48 tests, all passing, < 1 second

See CONTRIBUTING.md.

---

Roadmap

• Content-addressed object store (git-identical architecture)
• TOON format (compact line-oriented memory format)
• MCP server — Claude Code, Cursor, Windsurf, Cline, Continue.dev
• HTTP server — ChatGPT Custom Actions, Gemini function calling
• BM25 relevance search (load only what matters)
• memgit stats — measured token savings proof
• memgit squash — scale to 10k+ sessions
• memgit git push/pull — team sync via standard git
• Flat memories/ directory — grep/diff/blame your memories
• D3.js graph visualization of memory relationships
• memgit resume + SessionStart hook — sessions start with "where we left off"
• memgit gc — space reclamation (mark-and-sweep, lossless squash archive)
• Multi-agent write safety — store lock, auto-merge commits, memgit merge
• PyPI + Homebrew (tap) + npm published (v0.1.5)
• Chocolatey (not yet live on community.chocolatey.org)
• Interactive setup wizard (memgit setup)
• Smart memgit init (auto-detects tool, no path needed)
• VS Code extension (v0.1.5, Marketplace: code416-memgit.memgit)
• JetBrains plugin (Phase 3)
• Semantic search via embeddings (Phase 4)
• memgit.dev website (live)
• Memory compression / auto-summarization (Phase 5)
• Team access control + audit trail (Phase 5)
• Memory marketplace — share reusable context packs (Phase 6)

---

License

MIT — see LICENSE.