Give Cursor a memory your team shares.
Decisions made in Slack threads, in Claude or in last week's chat can reach the editor. Tell any AI once and Cursor knows.
Why Cursor needs a shared memory
Cursor's rules live in each project's files. Anything you decide somewhere else, like in a chat with Claude or ChatGPT, never reaches them.
- Across projects.One memory for every repository you open, not a copy of the rules in each one.
- Across apps.Something saved from Claude or ChatGPT is there when Cursor needs it.
- Across your team.A convention one person saves is one every teammate's Cursor can find, with its source.
Connect Cursor in about a minute
- Sign in to Contexel and open the Connect tab.
- Choose Cursor and copy the settings it shows you.
- Add them to your Cursor config file,
~/.cursor/mcp.json, and save. - Restart Cursor. Contexel shows up in its list of MCP servers.
The settings name your Contexel address, https://contexel.ai/mcp. Copy them from the console, because the exact version depends on how your account signs in.
Then just talk to it
There's nothing new to learn. Ask your AI to remember something, and ask about it later from any connected app.
You, in ClaudeRemember for the team: all new API endpoints go under /v2 and return JSON errors.
Later, a teammate in CursorAdd an endpoint that lists invoices.
Questions
Does this replace Cursor rules?
No. Keep your rules for things specific to one project. Use Contexel for what your team knows across projects and apps.
Does my code get uploaded?
No. Contexel only stores what you or your AI choose to save as a memory. It does not read your repository.
Who can see what I save?
You start with a private workspace that only you can use. You choose when to create a shared one and who to invite.
Using other AI apps too?
It's the same memory everywhere. Connect as many apps as you like.