Peer-to-peer AI memory sharing — approved on both sides.
Any tool · Any model · Any teammate · Your machine only
Every vendor builds memory in its own proprietary, closed format.
Spent 6 months "training" Claude or ChatGPT on your preferences? Want to switch? Start from zero.
Every AI memory product is per-user by design. There's no way to share context with a colleague or friend.
"Deleted" is a promise, not a guarantee. You don't know what's saved, where, or who has access.
Powerful tools exist. But every one of them stores memory per-account, in a format you cannot inspect, verify, or hand to anyone else.
| Tool | Cross-model portability | Self-hosted | User-to-user sharing | Consent flow |
|---|---|---|---|---|
| Claude / ChatGPT Memory | ✗ | ✗ | ✗ | ✗ |
| Mem0 | ✓ | ✓ | ✗ | ✗ |
| OMP (Open Memory Protocol) | ✓ | ✓ | ✗ | ✗ |
| Portable Memory (MacPaw) | ✓ | ✓ | ✗ | ✗ |
| memshare | ✓ | ✓ | ✓ | ✓ |
Make memory a file and three things follow — in this order.
Inspect it. Move it. Then share it, with approval on both sides.
1 Inspect — plain JSON. Greppable, diffable, deleted when you delete it.
2 Move — one store, every MCP tool, every machine you own.
3 Share — item by item, preview and approval on both sides.
✓ No central server. No vendor can revoke any of it.
Each person runs their server locally.
No central server sees everything.
Your memory is local JSON files. The MCP server is one adapter over them — anything else can be written the same way.
If MCP disappears tomorrow — your data is still here. Just files.
Three modes control what gets written down. What gets shared is gated separately, and always by you.
Claude saves what it learns as you work, choosing shareable for project facts and private for anything about you. Either way nothing leaves until you approve an export.
Claude proposes what to save at the end of a conversation. You approve, edit, or skip.
"Save this to memory" — only saved when you explicitly ask. Maximum control.
Auto mode — what actually happens, with no command and no prompt:
Starts local on your machine. Same code deploys to a team server when you're ready.
Not a real server — a local process that Claude starts and stops automatically. No open ports, no Docker, no cloud.
Human-readable, diffable, git-friendly. Sync between devices via Dropbox or git.
Selective export → preview → approval → import. Send the file however you want — email, Slack, AirDrop.
Every memory item is tagged private or shareable at creation time. Not all-or-nothing.
Sensitive PII (health, financial, government IDs) is blocked automatically before export. Conservative by default.
Before a file exists, Alice is shown the exact list — and what was held back, and why. She approves by saying so.
Bob sees every incoming item before any of it lands, and takes only what he wants. Nothing he already had is overwritten.
None of this is typed at a terminal. You talk to your AI; the context accumulates, and moves, on its own.
DAY 0 · DANA (TECH LEAD)
Install once. After that she never types a memshare command at all — she just works, and her AI records what it learns.
DAY 14 · DANA
~40 memories, already marked as they were learned. She checks, corrects anything by saying so, and asks for the ones Sam needs. She sees the list before anything is written.
DAY 15 · SAM JOINS
Sam looks before he takes anything, then accepts item by item. After that he is back to plain conversation — and his AI already knows the codebase.
Nobody wrote a wiki page, and nobody typed a command. Every step was a sentence — and both sides still saw exactly what was changing hands before it moved.
Dana joins the team. Instead of spending a week "teaching" her Claude about the project, the tech lead shares a bundle — internal terms, architecture decisions, code conventions.
Uri finished his part of the project. Instead of the next freelancer starting from zero, Uri shares a bundle — client preferences, decision history, technical gotchas.
The designer’s AI knows the component conventions. The backend devs’ AI knows the API contract. One bundle each way and neither side re-explains. Different people and different vendors — the only tool that crosses both.
A senior engineer "taught" their Claude about best practices, design patterns, and review guidelines. Instead of writing a guide — they share the memory directly.
Claude Code today, Cursor tomorrow, Windsurf next month. There is nothing to migrate — every MCP client reads the same local store. Same context, zero commands. (A ChatGPT route is planned.)
Long conversation with rich context. Open a new chat and have to explain everything again. With memshare — the MCP server auto-injects relevant context.
Claude for code, ChatGPT for research, Gemini for document analysis. All read from the same local memory store — no duplicating context three times.
Working on the laptop at home and the desktop at the office. The memory store syncs via git or Dropbox — because it's just JSON files.
Your memories are the same JSON files in all four, so moving between them costs you nothing — no migration, no re-setup, nothing left behind.
WE SUGGEST STARTING HERE
💻Files on your laptop.
MCP as local process.
Sharing via:
Export file → email/Slack → import file
PLANNED
📂Dropbox, Google Drive,
or a git repo.
Sharing via:
Auto-sync through the shared folder
PLANNED
🖥️Docker on your VPS.
Team connects remotely.
Sharing via:
By username on the server
PLANNED
☁️Managed infra.
Same protocol, zero ops.
Sharing via:
Username, invite link, or email
Your memory belongs to you.
You decide what to share, with whom, and for how long.