Cursor
Cursor memory between sessions
What Cursor's agent worked out in yesterday's session, such as which files matter and what broke, is gone in today's session unless something wrote it down.
Hippo is one place to write it down: a local store that Cursor's agent reads and writes through MCP, shared with Claude Code, Codex and your other agents. No account, no API key.
Set it up
$ npm install -g hippo-memory
$ hippo init hippo init adds its instructions to the project's AGENTS.md if it has one, and Cursor
reads that file from the project root. Then give Cursor's agent the memory tools by adding hippo to
.cursor/mcp.json:
{
"mcpServers": {
"hippo-memory": {
"command": "hippo",
"args": ["mcp"]
}
}
}
No global install? "command": "npx", "args": ["-y", "hippo-memory", "mcp"] works too. The agent gets
tools such as hippo_recall, hippo_remember and hippo_outcome; the
MCP server page lists all 13.
Bring your rules with you
# preview first, then import
$ hippo import --cursor .cursor/rules --dry-run
$ hippo import --cursor .cursor/rules hippo import --cursor reads the whole .cursor/rules tree, every .mdc and
.md file in it, and turns what is in them into memories, so lessons written for Cursor reach your
other agents too. It reads an older single .cursorrules file as well.
One store for every agent
Use Claude Code on Monday, Cursor on Tuesday and Codex on Wednesday: same .hippo/ store, same
memories. A lesson Claude Code stored after a failed build is there when Cursor's agent opens the project.
Cursor's rules are instructions you write: project rules in .cursor/rules, user rules, team rules and
AGENTS.md
(Cursor docs (opens in new tab),
checked 2026-09-28). Keep those for standing instructions. Cursor also had Memories, a beta in Cursor 1.0 that
remembered facts from conversations per project
(changelog (opens in new tab)).
Cursor staff say the feature was removed starting in version 2.1.x
(forum (opens in new tab)).
Hippo holds what the agent learns as it works, in a store your other agents share, and lets you mark a lesson
wrong with hippo outcome --bad or replace a fact that changed with hippo supersede.
How do I give Cursor memory between sessions?
hippo init adds its instructions to AGENTS.md if the project has one, and Cursor reads that file from the project root. The MCP server gives Cursor's agent tools to recall and store memories once you add hippo mcp to .cursor/mcp.json. Older hippo versions wrote the block to .cursorrules; hippo hook uninstall cursor removes it from there, and from AGENTS.md only when the block there is Cursor's own and unedited (hippo hook install cursor puts it back). A block written for Codex or another agent stays, since Cursor reads it too, and so does an edited block, since hippo cannot tell whose it is. hippo import --cursor .cursor/rules turns your existing rules into memories; it reads an older single .cursorrules file too.
Can I use hippo as an MCP memory server?
Yes. hippo mcp runs the server over stdio, and npx -y hippo-memory mcp runs it without a global install. Add it to the MCP config of Claude Desktop, Cursor, Windsurf (now Devin Desktop), Cline or any other client (example above); in Claude Code, run claude mcp add hippo-memory -- hippo mcp. The agent gets tools such as hippo_recall, hippo_remember and hippo_outcome.
Where does hippo keep my data?
On your machine, in SQLite: .hippo/hippo.db in each project, plus a global store in ~/.hippo/ for lessons shared across projects, with markdown mirrors you can read and commit. Recall makes no network call by default. Text goes to an outside provider only through features that use one: an API embedder, the Jev or LLM reranker, hippo refine, and the fact extraction hippo sleep runs through Anthropic's API whenever ANTHROPIC_API_KEY is set in its environment. To turn that last one off, set {"extraction":{"enabled":false}} in .hippo/config.json.