{"slug": "how-to-add-persistent-memory-to-cursor-with-mcp-2026", "title": "How to Add Persistent Memory to Cursor with MCP (2026)", "summary": "Edward, co-founder of Mnemoverse, explains three approaches to giving Cursor persistent memory: built-in Rules and Memories, local-first MCP memory servers like Basic Memory, and managed MCP memory services. He outlines tradeoffs for each, noting that local-first options keep data on-device while managed services offer portability across tools.", "body_md": "Cursor starts every session with amnesia: yesterday's architecture discussion, the migration you agreed not to touch, your preference for pnpm, all gone. There are three real ways to fix that. One, use Cursor's built-in Rules and Memories. Two, run a local-first MCP memory server such as Basic Memory, which keeps everything on your machine. Three, connect a managed memory service over MCP so the same memory follows you across tools. Disclosure up front: I'm Edward, co-founder of Mnemoverse, which is one of those managed services, so I'm biased about option three. The other two are genuinely the right choice for plenty of people, and I'll try to show where.\n\nRules are instruction files Cursor loads into the model's context. Project rules live in `.cursor/rules`\n\n, and there are global user rules in settings. They are memory in the \"sticky note\" sense: static text you write and curate by hand. That is exactly right for stable conventions. \"Use pnpm.\" \"Tests mirror the src layout.\" \"Never touch the legacy billing module.\" Write those once and stop repeating yourself.\n\nMemories is Cursor's automatic layer: the editor generates memories from your sessions so context carries forward. I have to be honest here: availability and behavior vary by version. The feature has moved around and changed between releases, so check what your build actually does before relying on it. When it is present and fits your workflow, it is the zero-setup option.\n\nThe limits are structural. Everything stays inside Cursor. Open Claude Code or VS Code and none of it comes with you. Rules do not learn from your sessions, and because Memories varies by version, what it captures is worth auditing rather than assuming.\n\nMCP (Model Context Protocol) lets you plug external tools into Cursor's agent, and memory servers are a natural fit. A local-first server keeps every byte on your machine.\n\nBasic Memory is a good example: an AGPL-3.0 MCP server that stores your memories as Markdown files on disk. That design decision does a lot of work. Your memory becomes a folder you can grep, edit in any editor, back up, and version in a private repo. Nothing leaves your machine, it works offline, there is no per-query cost, and there is no vendor to trust or outlive.\n\nThe tradeoffs are equally clear. It lives on one machine unless you sync it yourself. Retrieval tends to be simpler than in hosted systems, with no learned ranking or feedback loop. And maintenance is yours. But if you operate under compliance constraints, or you simply do not want project context on anyone's server, this option wins and it is not close.\n\nThis is what I build, so read this section with that bias in mind. The point of a managed service is not that it is smarter than your local setup. The point is portability plus mechanics you do not have to build yourself. One API key or OAuth works across Claude Code, Cursor, VS Code, and ChatGPT (ChatGPT connects via a Custom GPT action), so the fact your agent learned in Cursor on Monday is available in Claude Code on Tuesday.\n\nSetup takes a few minutes:\n\nStep 1. Sign up at [console.mnemoverse.com](https://console.mnemoverse.com). The free tier is 1,000 queries per day and 10,000 atoms, no credit card.\n\nStep 2. Copy your API key.\n\nStep 3. Add the server to `.cursor/mcp.json`\n\nin your project (or the global `~/.cursor/mcp.json`\n\n):\n\n```\n   {\n     \"mcpServers\": {\n       \"mnemoverse\": {\n         \"command\": \"npx\",\n         \"args\": [\"-y\", \"@mnemoverse/mcp-memory-server@latest\"],\n         \"env\": {\n           \"MNEMOVERSE_API_KEY\": \"mk_live_YOUR_KEY\",\n           \"MNEMOVERSE_API_URL\": \"https://core.mnemoverse.com/api/v1\"\n         }\n       }\n     }\n   }\n```\n\nStep 4. Reload MCP servers in Cursor's settings. The agent now sees twelve tools: seven core memory tools (`memory_write`\n\n, `memory_read`\n\n, `memory_feedback`\n\n, `memory_stats`\n\n, `memory_list_recent`\n\n, `memory_delete`\n\n, `memory_delete_domain`\n\n), four Beta tools for shared rooms (create, invite, join, list), and a vault listing tool.\n\nIf you would rather not manage a key at all, there is a remote connector at `mnemoverse.com/mcp`\n\nthat signs you in through OAuth in the browser. It exposes ten of those twelve; the two delete tools are npx-only, which is the reason to prefer the config above.\n\nWhat actually happens under the hood:\n\n`memory_read`\n\nreturns atoms ranked using those signals: importance, the learned associations, and past feedback.`memory_feedback`\n\nwith a score from -1 to 1 when a recalled memory helped or misled it, and that signal re-ranks future recall.These mechanics are documented at [mnemoverse.com/docs](https://mnemoverse.com/docs).\n\nOpenness, stated plainly: the MCP server and the Python SDK are MIT licensed, but the engine is hosted and not open source. Your atoms live on our servers. That is the real tradeoff against option two, and `memory_delete`\n\n/ `memory_delete_domain`\n\nexist precisely so you stay in control of what is kept. Pro is $29/month if you outgrow the free tier. Docs are at [mnemoverse.com/docs](https://mnemoverse.com/docs).\n\nIf Cursor is your only tool and your needs are conventions plus light session carry-over, use the native features. It is the least machinery, and Rules in particular are underrated.\n\nIf privacy, latency, or data ownership dominates, go local-first. Basic Memory if Markdown on disk appeals to you; the pattern matters more than the specific server.\n\nIf you work across several tools, managed memory is the category to look at, and I would compare honestly within it. Mem0 offers an Apache-2.0 open-source SDK plus a managed cloud and focuses on extracting facts from conversations. Zep builds on the open-source Graphiti temporal knowledge graph, where facts carry validity windows, a real advantage if you care about what was true when. If your stack is LangGraph, LangMem is native to its Long-term Memory Store and is the path of least resistance. Cognee is the one to evaluate for a self-hosted knowledge graph. Letta gives you a whole agent framework with self-editing memory rather than a bolt-on. Supermemory leans on connectors (Notion, Google Drive, Gmail, S3) if your context lives in documents more than code. Mnemoverse's angle is the cross-tool key plus the learning mechanics above.\n\nOn performance claims: vendor-reported numbers are contested, ours included; run your own evals on your own recall tasks before you commit.\n\n**Does MCP memory slow Cursor down?**\n\nEvery read or write is a tool call the agent chooses to make, so the cost shows up a few times per session, not per keystroke. A local server answers on localhost; a managed one adds a network hop. If that hop bothers you, that is a point for option two.\n\n**Does a memory service see my whole codebase?**\n\nNo. It only receives what the agent explicitly sends through the memory tools: the memories it writes and the queries it reads with. You can inspect and delete everything: with Mnemoverse via the delete tools, with Basic Memory by deleting a Markdown file. If \"only what the tools send\" is still too much for your threat model, that is option two's whole argument, and it is a fair one.\n\n**Can other tools share the memory Cursor builds?**\n\nNative Rules and Memories stay in Cursor. A local MCP server can be mounted by any MCP client on the same machine. A managed service shares across machines and tools; with Mnemoverse, one API key or OAuth covers Claude Code, Cursor, VS Code, and ChatGPT, the last via a Custom GPT action.\n\nI build one of these options, so weigh my framing accordingly. And if you try any of the three and hit something rough, tell me in the comments. I read them.", "url": "https://wpnews.pro/news/how-to-add-persistent-memory-to-cursor-with-mcp-2026", "canonical_source": "https://dev.to/izgorodin/how-to-add-persistent-memory-to-cursor-with-mcp-2026-148n", "published_at": "2026-08-12 12:06:00+00:00", "updated_at": "2026-08-12 12:17:43.719856+00:00", "lang": "en", "topics": ["developer-tools", "ai-tools", "ai-agents"], "entities": ["Cursor", "Mnemoverse", "Basic Memory", "MCP", "Claude Code", "VS Code", "ChatGPT"], "alternates": {"html": "https://wpnews.pro/news/how-to-add-persistent-memory-to-cursor-with-mcp-2026", "markdown": "https://wpnews.pro/news/how-to-add-persistent-memory-to-cursor-with-mcp-2026.md", "text": "https://wpnews.pro/news/how-to-add-persistent-memory-to-cursor-with-mcp-2026.txt", "jsonld": "https://wpnews.pro/news/how-to-add-persistent-memory-to-cursor-with-mcp-2026.jsonld"}}