pūrmemo
Provides persistent memory for AI assistants like Claude, enabling them to remember user identity, projects, and conversations across sessions and platforms via natural language commands.
README
pūrmemo
Memory for your AI tools. Claude remembers who you are, what you're working on, and what you said last time — across every session, on every platform.
Just want it in ChatGPT or Claude.ai in your browser? Get the Chrome Extension instead.
Install in 30 seconds
1. Paste this into your terminal
Mac or Linux:
curl -fsSL https://app.purmemo.ai/install | sh
Windows (PowerShell):
irm https://app.purmemo.ai/install.ps1 | iex
That's it for installing. The script handles everything — it'll install Node if you don't have it, set up the purmemo command, and tell you when it's done.
2. Type purmemo
purmemo
Your browser opens. Sign in (or create a free account). Close the tab when it says you're done.
3. Restart Claude
Quit and reopen Claude Desktop (or Claude Code). pūrmemo is now connected.
You'll know it worked when your next Claude session opens with a header like:
pūrmemo v15.7.20 · you@example.com · Free · 0 memories
Try saying "save this" at the end of a conversation, or "what was I working on?" at the start of a new one.
If something goes wrong
purmemo command not found? Close and reopen your terminal, then try again. New commands sometimes need a fresh shell.
You see "Failed to read token" or "bad decrypt"? Run:
purmemo --update && purmemo init
This clears any stale credentials and signs you back in. (Fixed in v15.7.20+ — if you're on an older version, this is a one-time thing.)
Anything else? Open an issue at github.com/purmemo-ai/purmemo-mcp/issues — we read every one.
What it does
- Remembers everything — save any conversation, recall it later by typing what you remember about it.
- Knows who you are — your role, your projects, your stack — loaded automatically into every new session.
- Works everywhere — Claude Code, Claude Desktop, Cursor, Windsurf, Zed, anything that speaks MCP.
Three slash commands you'll use most:
| You type | What happens |
|---|---|
/save |
Saves this conversation. Use the same title later → updates the same memory. |
/recall <topic> |
Search your memories in plain English. |
/context |
At the start of a session — loads who you are and what you were last working on. |
Other ways to install
<details> <summary><b>I already have Node.js</b></summary>
npm install -g purmemo-mcp && purmemo
Or, run it once without installing globally:
npx purmemo-mcp@latest init
</details>
<details> <summary><b>Claude Desktop — hosted (recommended, no setup)</b></summary>
Open Claude Desktop → Settings → Developer → Edit Config and add:
{
"mcpServers": {
"purmemo": {
"url": "https://mcp.purmemo.ai/mcp/messages",
"transport": "streamable-http"
}
}
}
Restart Claude Desktop. You'll be prompted to sign in via OAuth on first use.
</details>
<details> <summary><b>Claude Desktop — local (advanced)</b></summary>
Get your API key from app.purmemo.ai → Settings → API Keys, then edit:
- macOS:
~/Library/Application Support/Claude/claude_desktop_config.json - Windows:
%APPDATA%\Claude\claude_desktop_config.json
{
"mcpServers": {
"purmemo": {
"command": "npx",
"args": ["-y", "purmemo-mcp"],
"env": { "PURMEMO_API_KEY": "your-api-key-here" }
}
}
}
Restart Claude Desktop after saving.
</details>
<details> <summary><b>Cursor / Windsurf / Zed</b></summary>
Cursor — edit ~/.cursor/mcp.json:
{ "mcpServers": { "purmemo": { "command": "npx", "args": ["-y", "purmemo-mcp"], "env": { "PURMEMO_API_KEY": "your-api-key" } } } }
Windsurf — edit ~/.codeium/windsurf/mcp_config.json (same shape as above).
Zed — edit ~/.config/zed/settings.json, add under context_servers:
{ "context_servers": { "purmemo": { "command": { "path": "npx", "args": ["-y", "purmemo-mcp"], "env": { "PURMEMO_API_KEY": "your-api-key" } } } } }
</details>
Pricing
| Plan | Price | Recalls | Saves |
|---|---|---|---|
| Free | $0 | 50/month | Unlimited |
| Pro | $19/month | Unlimited | Unlimited |
For developers
Looking for the technical stuff? It's all here:
- Tools, resources, prompts reference — every MCP tool the server exposes (
save_conversation,recall_memories,commit,snapshot,run_workflow, etc.) - Living document semantics —
mode='append'vsmode='replace', chunking behavior, ADR-036/038 details. - Identity layer — the cognitive fingerprint that loads into every session.
- Architecture decisions (ADRs) — every design decision, with context and trade-offs.
- Source for the install scripts — read before running, if you want.
Links
- Dashboard — view and manage memories
- Chrome Extension — for ChatGPT, Claude.ai, Gemini in browser
- Privacy Policy — encrypted in transit and at rest, never shared
- Support / Issues
License
The MCP connector code in this repo is MIT licensed. The pūrmemo platform, API, and backend are proprietary.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.