Cursor memory setup
Cursor Rules are instructions humans write — .cursor/rules files and dashboard-managed Team Rules that guide agent behavior. They are the right place for conventions and standards, and Cursor documentation describes Rules as its persistence mechanism. What Rules do not do is update themselves with what agents discover while working. StremAI adds that learned layer over MCP: agents store lessons as they go, and connected agents can recall them in later sessions — in Cursor, and in other tools connected to the same memory. Rules remain the place for "always do X"; StremAI holds "here is what we learned."
Verified with Claude Code and Claude Desktop; Cursor, Codex, Windsurf, OpenClaw, and other MCP clients are supported or in active verification.
Markdown versionHow Cursor and StremAI fit together
Cursor Rules remain the place for standards and conventions. StremAI holds learned context such as pitfalls, decisions, and project notes that connected agents can recall later.
Use the hosted MCP endpoint with OAuth/browser sign-in when Cursor supports it, or use the API-key fallback for headless or blocked environments.
Verify the connection
Store a memory about staging, start a new chat, and ask Cursor what it knows about staging. A working connection should recall the stored entry.
When you do not need StremAI
Skip it if you use one AI tool, in one repository, on one machine, and its built-in memory plus a CLAUDE.md or AGENTS.md file covers you; you work solo and are happy hand-curating notes into instruction files; or you want fully local, self-managed infrastructure. Open-source MCP memory servers are a reasonable choice if you prefer running your own. StremAI earns its place when agents span multiple tools, machines, or teammates, and when you want memory that is shared, attributed, and user-controlled without operating the layer yourself.
FAQ
How do I verify Cursor is using StremAI memory?
Store a memory: "our staging database resets every Sunday." Start a new session and ask what the agent knows about staging. A connected agent should recall the stored entry rather than asking you to repeat it.
Does StremAI replace Cursor instructions?
No. Keep instruction files and rules for commands, conventions, and policy. StremAI adds the learned memory layer: decisions, pitfalls, and context discovered while agents work.
What is the markdown version of this page?
Use https://stremai.com/docs/cursor-memory.md when an agent needs a plain-text copy without navigation chrome.
Start with a real connection
OAuth/browser sign-in is preferred. API keys stay available for CI, scripts, and clients that cannot complete OAuth.