### Connect to Supermemory MCP via SSE

Source: https://context7.com/supermemoryai/supermemory-mcp/llms.txt

Establishes a persistent connection to the Supermemory MCP server using Server-Sent Events. Includes commands for direct connection and automated client installation for tools like Claude, Cursor, and Cline.

```bash
# Connect to the MCP SSE endpoint with your user ID
curl -N "https://mcp.supermemory.ai/{userId}/sse"

# Or with a session ID for resuming connections
curl -N "https://mcp.supermemory.ai/{userId}/sse?sessionId={sessionId}"

# One-click installation for various MCP clients
npx install-mcp i https://mcp.supermemory.ai/{userId}/sse --client claude
npx install-mcp i https://mcp.supermemory.ai/{userId}/sse --client cursor
npx install-mcp i https://mcp.supermemory.ai/{userId}/sse --client cline
npx install-mcp i https://mcp.supermemory.ai/{userId}/sse --client windsurf
```

--------------------------------

### GET /sse

Source: https://context7.com/supermemoryai/supermemory-mcp/llms.txt

Establishes a Server-Sent Events (SSE) connection for real-time communication with the MCP server.

```APIDOC
## GET /{userId}/sse

### Description
Establishes a persistent SSE connection for MCP communication, enabling real-time memory operations.

### Method
GET

### Endpoint
https://mcp.supermemory.ai/{userId}/sse

### Parameters
#### Path Parameters
- **userId** (string) - Required - Unique identifier for the user session.

#### Query Parameters
- **sessionId** (string) - Optional - Session ID for resuming existing connections.

### Response
#### Success Response (200)
- **Connection** (EventStream) - Persistent SSE stream for MCP protocol messages.

#### Response Example
{ "status": "connected", "protocol": "mcp-sse" }
```

--------------------------------

### Self-Hosting Environment Configuration

Source: https://context7.com/supermemoryai/supermemory-mcp/llms.txt

Configuration template for setting up the project in a self-hosted environment using environment variables.

```bash
SUPERMEMORY_API_KEY=your_supermemory_api_key_here
AUTH_SECRET=your_session_secret_here
```

--------------------------------

### Execute Development and Deployment Commands

Source: https://context7.com/supermemoryai/supermemory-mcp/llms.txt

A set of CLI commands used to manage the project lifecycle. These commands handle local development, production builds, deployment to Cloudflare, and type generation.

```bash
bun run dev          # Start development server
bun run build        # Build for production
bun run deploy       # Build and deploy to Cloudflare
bun run cf-typegen   # Generate Cloudflare types
```

--------------------------------

### POST /add

Source: https://context7.com/supermemoryai/supermemory-mcp/llms.txt

Stores user information, preferences, and behaviors into the memory system.

```APIDOC
## POST /add

### Description
Stores user information, preferences, and behaviors. This tool is invoked when detecting significant user traits or patterns.

### Method
POST

### Request Body
- **thingToRemember** (string) - Required - The information to store in memory.

### Request Example
{
  "thingToRemember": "User prefers TypeScript over JavaScript"
}

### Response
#### Success Response (200)
- **type** (string) - Response type (text).
- **text** (string) - Confirmation message.

#### Response Example
{
  "type": "text",
  "text": "Memory added successfully"
}
```

--------------------------------

### POST /?index (Fetch Memories)

Source: https://context7.com/supermemoryai/supermemory-mcp/llms.txt

Retrieves the list of memories associated with a specific user ID.

```APIDOC
## POST /?index

### Description
Lists all memories for a given user through the web interface action handler.

### Method
POST

### Endpoint
/?index

### Request Body
- **userId** (string) - Required - The unique identifier for the user.
- **action** (string) - Required - Must be set to "fetch".

### Request Example
{
  "userId": "abc123xyz",
  "action": "fetch"
}

### Response
#### Success Response (200)
- **success** (boolean) - Status of the request.
- **memories** (array) - List of memory objects.

#### Response Example
{
  "success": true,
  "memories": [
    {
      "id": "mem_abc123",
      "title": "User prefers TypeScript",
      "summary": "User prefers TypeScript over JavaScript",
      "containerTags": ["abc123xyz"]
    }
  ]
}
```

--------------------------------

### POST /search

Source: https://context7.com/supermemoryai/supermemory-mcp/llms.txt

Performs semantic search across user memories using vector-based matching.

```APIDOC
## POST /search

### Description
Performs semantic search across user memories and patterns to retrieve relevant context.

### Method
POST

### Request Body
- **informationToGet** (string) - Required - The query to search for in memories.

### Request Example
{
  "informationToGet": "What programming languages does the user prefer?"
}

### Response
#### Success Response (200)
- **type** (string) - Response type (text).
- **text** (string) - The retrieved memory content.

#### Response Example
{
  "type": "text",
  "text": "User prefers TypeScript over JavaScript"
}
```

--------------------------------

### Configure Cloudflare Worker with wrangler.jsonc

Source: https://context7.com/supermemoryai/supermemory-mcp/llms.txt

This configuration file defines the Cloudflare Worker settings, including compatibility dates, entry points, and Durable Object bindings. It enables Node.js compatibility and sets up database migrations for persistent storage.

```jsonc
{
  "name": "supermemory-mcp-app",
  "compatibility_date": "2025-04-04",
  "main": "./workers/app.ts",
  "compatibility_flags": ["nodejs_compat", "nodejs_compat_populate_process_env"],
  "durable_objects": {
    "bindings": [{ "name": "MY_DO", "class_name": "MyDurableObject" }]
  },
  "migrations": [{ "tag": "v1", "new_sqlite_classes": ["MyDurableObject"] }]
}
```

--------------------------------

### Configure Supermemory System Prompt

Source: https://context7.com/supermemoryai/supermemory-mcp/llms.txt

Defines the system prompt schema used to instruct LLMs on how to proactively manage user memory. This ensures consistent storage and retrieval behavior during sessions.

```typescript
{
  "name": "Supermemory Prompt",
  "description": "A prompt that gives information about supermemory.",
  "completion": ["supermemory", "memory", "supermemory api"]
}

// Response format when prompt is invoked:
{
  "role": "user",
  "content": {
    "type": "text",
    "text": "IMPORTANT: You MUST use Supermemory tools proactively to be an effective assistant..."
  }
}
```

--------------------------------

### Implement addToSupermemory Tool

Source: https://context7.com/supermemoryai/supermemory-mcp/llms.txt

Defines the schema and interaction pattern for the addToSupermemory tool. This tool enables LLMs to persist user preferences and traits into the memory store.

```typescript
{
  "name": "addToSupermemory",
  "description": "Store user information, preferences, and behaviors. Run on explicit commands ('remember this') or implicitly when detecting significant user traits, preferences, or patterns.",
  "inputSchema": {
    "type": "object",
    "properties": {
      "thingToRemember": {
        "type": "string",
        "description": "The information to store in memory"
      }
    },
    "required": ["thingToRemember"]
  }
}

// Example POST request
{
  "thingToRemember": "User prefers TypeScript over JavaScript and uses Tailwind CSS for styling"
}
```

--------------------------------

### Implement searchSupermemory Tool

Source: https://context7.com/supermemoryai/supermemory-mcp/llms.txt

Defines the schema and interaction pattern for the searchSupermemory tool. It utilizes semantic vector matching to retrieve relevant historical context based on user queries.

```typescript
{
  "name": "searchSupermemory",
  "description": "Search user memories and patterns. Run when explicitly asked or when context about user's past choices would be helpful. Uses semantic matching to find relevant details across related experiences.",
  "inputSchema": {
    "type": "object",
    "properties": {
      "informationToGet": {
        "type": "string",
        "description": "The query to search for in memories"
      }
    },
    "required": ["informationToGet"]
  }
}

// Example POST request
{
  "informationToGet": "What programming languages does the user prefer?"
}
```

--------------------------------

### Restore Session and Parse MCP URL

Source: https://context7.com/supermemoryai/supermemory-mcp/llms.txt

Restores a user session and demonstrates how to extract a user ID from an MCP SSE URL using regex.

```typescript
const formData = new FormData();
formData.append("userId", "previousUserId123");
formData.append("action", "restore");

const response = await fetch("/?index", {
  method: "POST",
  body: formData
});

const mcpUrl = "https://mcp.supermemory.ai/abc123xyz/sse";
const urlPattern = /\/([^\/]+)\/sse/;
const match = mcpUrl.match(urlPattern);
const extractedUserId = match[1];
```

--------------------------------

### POST /?index (Update Memory)

Source: https://context7.com/supermemoryai/supermemory-mcp/llms.txt

Updates the content of an existing memory.

```APIDOC
## POST /?index

### Description
Updates the content of an existing memory through the web interface.

### Method
POST

### Endpoint
/?index

### Request Body
- **userId** (string) - Required - The user identifier.
- **action** (string) - Required - Must be set to "update".
- **memoryId** (string) - Required - The ID of the memory to update.
- **content** (string) - Required - The new content for the memory.

### Request Example
{
  "userId": "abc123xyz",
  "action": "update",
  "memoryId": "mem_abc123",
  "content": "Updated content text"
}

### Response
#### Success Response (200)
- **success** (boolean) - Status of the request.
- **message** (string) - Confirmation message.

#### Response Example
{
  "success": true,
  "message": "Memory updated successfully"
}
```

--------------------------------

### POST /?index (Delete Memory)

Source: https://context7.com/supermemoryai/supermemory-mcp/llms.txt

Deletes a specific memory entry by its ID.

```APIDOC
## POST /?index

### Description
Removes a specific memory by its ID through the web interface.

### Method
POST

### Endpoint
/?index

### Request Body
- **userId** (string) - Required - The user identifier.
- **action** (string) - Required - Must be set to "delete".
- **memoryId** (string) - Required - The ID of the memory to delete.

### Request Example
{
  "userId": "abc123xyz",
  "action": "delete",
  "memoryId": "mem_abc123"
}

### Response
#### Success Response (200)
- **success** (boolean) - Status of the request.
- **message** (string) - Confirmation message.

#### Response Example
{
  "success": true,
  "message": "Memory deleted successfully"
}
```

--------------------------------

### Fetch User Memories via Web API

Source: https://context7.com/supermemoryai/supermemory-mcp/llms.txt

Retrieves a list of memories for a specific user. It sends a POST request with the user ID and action type to the web interface.

```typescript
const formData = new FormData();
formData.append("userId", "abc123xyz");
formData.append("action", "fetch");

const response = await fetch("/?index", {
  method: "POST",
  body: formData
});
```

--------------------------------

### Cloudflare Durable Object Session Management

Source: https://context7.com/supermemoryai/supermemory-mcp/llms.txt

Implements persistent SSE connections and session routing using Cloudflare Durable Objects. This ensures state is maintained across multiple requests.

```typescript
export class MyDurableObject extends DurableObject<Env> {
  transport?: SSEHonoTransport;

  async fetch(request: Request) {
    const url = new URL(request.url);
    const userId = url.pathname.split("/")[1];

    if (!this.transport) {
      this.transport = new SSEHonoTransport(`/${userId}/messages`, this.ctx.id.toString());
    }
  }
}

export default {
  async fetch(request: Request, env: Env, ctx: ExecutionContext) {
    const url = new URL(request.url);
    if (url.pathname.includes("sse") || url.pathname.endsWith("/messages")) {
      const sessionId = url.searchParams.get("sessionId");
      const namespace = env.MY_DO;
      const stub = sessionId ? namespace.get(namespace.idFromString(sessionId)) : namespace.get(namespace.newUniqueId());
      return stub.fetch(request);
    }
    return requestHandler(request, { cloudflare: { env, ctx } });
  }
};
```

--------------------------------

### Update Memory via Web API

Source: https://context7.com/supermemoryai/supermemory-mcp/llms.txt

Updates the content of an existing memory. The request includes the user ID, memory ID, and the new content string.

```typescript
const formData = new FormData();
formData.append("userId", "abc123xyz");
formData.append("action", "update");
formData.append("memoryId", "mem_abc123");
formData.append("content", "User strongly prefers TypeScript with strict mode enabled");

const response = await fetch("/?index", {
  method: "POST",
  body: formData
});
```

--------------------------------

### Delete Memory via Web API

Source: https://context7.com/supermemoryai/supermemory-mcp/llms.txt

Removes a specific memory entry by its unique ID. Requires the user ID and the target memory ID to perform the deletion.

```typescript
const formData = new FormData();
formData.append("userId", "abc123xyz");
formData.append("action", "delete");
formData.append("memoryId", "mem_abc123");

const response = await fetch("/?index", {
  method: "POST",
  body: formData
});
```

=== COMPLETE CONTENT === This response contains all available snippets from this library. No additional content exists. Do not make further requests.