Memory MCP Server
OfficialGive your AI persistent memory that survives across conversations
The official Memory MCP server implements a knowledge graph that your AI assistant can read and write across conversations. By default, Claude's memory resets after every conversation. With the Memory server, you can instruct it to remember facts, preferences, projects, and context that persist indefinitely. The knowledge is stored as a local JSON file on your machine — nothing goes to a cloud service. Essential for developers who want a consistent AI assistant that knows their preferences.
Install Config
Config file: ~/Library/Application Support/Claude/claude_desktop_config.json· Windows: %APPDATA%\Claude\claude_desktop_config.json
{
"mcpServers": {
"memory": {
"command": "npx",
"args": [
"-y",
"@modelcontextprotocol/server-memory"
]
}
}
}Available Tools (8)
| Tool name | Description |
|---|---|
| create_entities | Add new entities (people, projects, concepts) to memory |
| create_relations | Create relationships between entities |
| add_observations | Add facts or observations about an entity |
| delete_entities | Remove entities from the knowledge graph |
| delete_observations | Remove specific observations |
| delete_relations | Remove relationships between entities |
| read_graph | Read the entire knowledge graph |
| search_nodes | Search for specific nodes in the graph |
Example Prompts
- “Remember that I prefer TypeScript over JavaScript”
- “What do you know about my current project?”
- “Remember that I'm working on a Next.js app called QuickUtil”
- “Forget everything you know about my old job”
- “What are my coding preferences you've stored?”
Pros & Cons
Pros
- Persistent memory across all conversations
- All data stored locally — never sent to a server
- Structured knowledge graph — not just raw text
Cons
- Memory must be explicitly instructed — AI won't remember automatically
- Knowledge graph can become large and unwieldy over time
Who Is This For?
- ·Power users who want a consistent AI assistant that knows their preferences
- ·Developers who want AI to remember project context across sessions
When to use: Use when you're frustrated by having to re-explain your preferences and context at the start of every conversation. Set it up and instruct the AI to remember important facts about you and your work.
Supported Clients
Related Use Cases
Related Servers
Filesystem MCP Server
OfficialGive your AI read and write access to your local files
The official Filesystem MCP server from Anthropic provides secure, configurable access to your local file system. You sp…
Sequential Thinking MCP Server
OfficialEnable structured, multi-step reasoning for complex AI problems
The Sequential Thinking MCP server provides a framework for breaking complex problems into structured, sequential though…