Make your AI agents
remember your work
Stop re-explaining your codebase, your preferences, and yesterday's work. agimem is a shared memory layer your agents read and write over the Model Context Protocol (MCP) — so context survives the session, the project, and the tool.
Free to use. No credit card required. Setup takes under a minute.
Copy & paste into your agent to get started
Set up agimem as my memory server: https://agimem.dev/setup
Compatible with
Your agent remembers what didn't work.
Agents don't just read memory — they write to it. So the next agent, in any tool, never repeats yesterday's mistake.
Day 1 · Cursor · CI failure
Cursor tried pnpm install --frozen-lockfile in CI. Build broke (workspace catalog mismatch). Cursor fixed it with --no-frozen-lockfile and wrote a note back to the Capsule: “avoid --frozen-lockfile — breaks our pnpm catalog. use --no-frozen-lockfile.”
Without agimem
Day 2 · Claude
pnpm install --frozen-lockfile and then builds.With agimem
Day 2 · Claude
ci.pnpm.frozen_lockfile = avoid — breaks our catalogGot it — adding the job with pnpm install --no-frozen-lockfile so we don't hit yesterday's catalog issue.A static AGENTS.md can't learn. agimem lets every agent record what it figured out — so the next one, in any tool, starts smarter than the last.
Memory that lives outside the repo
AGENTS.md, CLAUDE.md, and .cursor/rules are great for one repo and one tool. agimem is the layer above them — shared across agents, projects, and machines.
Repo-local memory files
Stuck inside one repo, one tool
— Cursor reads .cursor/rules. Claude reads CLAUDE.md. Each tool, its own file.— A new project? Start the docs from scratch.— Multiple agents on the same task can't share state — only what's committed to git.— Personal preferences end up in every repo, or in none.
With agimem
One memory layer, every agent
Cross-agent — Claude, Cursor, Windsurf, and your custom MCP agents read and write the same Capsule.Cross-project — keep a personal Capsule for your preferences and reuse it everywhere.Cross-session — agents update memory as they work, so the next run picks up where the last left off.Cross-machine — nothing to commit, nothing to sync. It's just there.
Keep your AGENTS.md for repo-specific rules. Use agimem for everything that should outlive a single repo or agent.
Three steps to persistent memory
From zero to agent memory in under a minute.
Step 01
Create a Capsule
A Capsule is an isolated key-value store. Create one per agent, project, or use case.
Step 02
Generate an API key
Each Capsule gets its own API key. Share it only with the agent that needs access.
Step 03
Connect your agent
Point any MCP-compatible client at the capsule with your key. Agents can now remember things.
Show me the raw MCP config
▾
Show me the raw MCP config
mcp-config.json
Memory that follows your agents
Cursor saves a preference on Monday. Claude uses it on Wednesday. Windsurf avoids a bug on Friday. One shared Capsule, zero re-explaining.
A dashboard built for managing agent memory
Browse Capsules, rotate API keys, and inspect what your agents are actually remembering — all from one place.
Ready to try it?
Set up your first agent memory in under a minute.
Free to use. No credit card needed.
Everything agents need to remember
A minimal, secure, and fast memory layer built on open standards.
Isolation
Isolated Capsules
One Capsule per agent or project — each one is a fully separate memory space, no leaks across boundaries.
Sharing
Shared Memories
Point multiple agents at the same Capsule — they read and write the same context so they can collaborate without copy-pasting state.
Access
Simple Access Control
One API key per Capsule — share it with as many agents as you like, rotate it, or revoke it any time from the dashboard.
Compatibility
Works with Any MCP Client
Built on the open Model Context Protocol — supported today by Claude, Cursor, Windsurf, VS Code, and any custom MCP agent.
Built for the way agents actually work
Whether you're a solo developer or a team running fleets of agents, persistent memory means fewer repeated prompts and cleaner handoffs.
Coding conventions
Your style, every repo
Project structure, naming rules, and preferred libraries live in memory so your agent starts with your conventions instead of guessing.
e.g. “Always use Drizzle, never Prisma. Server actions live in lib/.”
Project onboarding
Drop into any codebase
Architecture notes, tech debt, and environment quirks live in memory so a new agent becomes useful quickly instead of rediscovering context.
e.g. “Monorepo with pnpm workspaces. Auth in Clerk. Deploy on Vercel.”
Personal preferences
Tone & formatting that stick
Save tone, formatting rules, and writing preferences once, then reuse them across every session without repeating the same setup prompt.
e.g. “Concise, no emojis, prefer bullets over paragraphs.”
Automation state
Resume where the job died
Job state, checkpoints, and outputs persist between runs so long-running automations resume cleanly instead of starting over.
e.g. “Last row processed: 4,281. Retry queue: empty.”
Questions developers ask before signing up
Everything you need to decide fast and get your first agent connected.
What is an MCP memory server?
An MCP memory server lets AI agents store and retrieve persistent context through the Model Context Protocol. Instead of resetting each session, agents can remember decisions, conventions, and project state.
Which AI tools can I use with agimem?
agimem works with MCP-compatible clients including Claude, Cursor, Windsurf, and other tools that support remote MCP servers.
How quickly can I set it up?
Most users can connect their first agent in under a minute. Create a Capsule, generate an API key, and paste one setup prompt.
Can multiple agents share memory?
Yes. You can point multiple agents to the same Capsule so they share context, or isolate each agent with separate Capsules for stricter boundaries.
Is agimem free to get started?
Yes. You can sign up and create your first Capsule without a credit card.