Skydeckai MCP Server Rememberizer

🚀 MCP Get Community Server - Rememberizer

A Model Context Protocol server for interacting with Rememberizer's document and knowledge management APIs. This server enables large language models to search, retrieve, and manage documents and integrations.

⚠️ Important Note

Please note that mcp-server-rememberizer is currently in development, and its features may change.

✨ Features

Resources

The server provides two types of resources: documents or Slack discussions.

Tools

1. rememberizer_search

   • Search documents based on semantic similarity.
   • Input:
     • q (string): A sentence of no more than 400 words for finding semantically similar paragraphs.
     • n (integer, optional): The number of similar documents to return (default: 5).
     • from (string, optional): An ISO 8601 formatted date with timezone (e.g., 2023-01-01T00:00:00Z). Use this parameter to filter results by a specific start date (default: none).
     • to (string, optional): An ISO 8601 formatted end date with timezone (e.g., 2024-01-01T00:00:00Z). Use this parameter to filter results by a specific end date (default: none).
   • Return: Text output of the search results.

2. rememberizer_agentic_search

   • Semantic similarity document search accelerated with an LLM Agent.
   • Input:
     • query (string): A sentence of no more than 400 words for finding semantically similar paragraphs. This query can be enhanced by our LLM Agent for better results.
     • n_chunks (integer, optional): The number of similar documents to return (default: 5).
     • user_context (string, optional): Additional context for the query. To get better context-aware results, it may be necessary to summarize the conversation so far (default: none).
     • from (string, optional): An ISO 8601 formatted start date with timezone (e.g., 2023-01-01T00:00:00Z). Use this parameter to filter results by a specific start date (default: none).
     • to (string, optional): An ISO 8601 formatted end date with timezone (e.g., 2024-01-01T00:00:00Z). Use this parameter to filter results by a specific end date (default: none).
   • Return: Text output of the search results.

3. rememberizer_get_chats

   • Retrieve user chat history.
   • Input: None
   • Return: A list of the user's chat history.

4. rememberizer_get_messages

   • Get messages in a specific chat.
   • Input:
     • chat_id (string): The ID of the chat to retrieve messages from.
   • Return: A list of messages in the specified chat.

5. rememberizer_add_message

   • Add a message to a specific chat.
   • Input:
     • chat_id (string): The ID of the chat to add the message to.
     • message (string): The content of the message to add.
   • Return: Information about the chat after the message is successfully added.

6. rememberizer_create_chat

   • Create a new chat record.
   • Input: None
   • Return: Information about the newly created chat record.

7. rememberizer_delete_chat

   • Delete a specified chat record.
   • Input:
     • chat_id (string): The ID of the chat to delete.
   • Return: The status after successful deletion.

8. rememberizer_clear_context

   • Clear the context.
   • Input: None
   • Return: The status after successful clearing.

📦 Installation

Via Smithery

npx smithery add mcp-server-rememberizer

Using the uv tool

1. Install dependencies:

npm install

2. Start the server:

node src/mcp_server_rememberizer.js

📚 Configuration

Set the following environment variables to configure the behavior of the MCP server:

• PORT: Specify the port on which the server runs (default: 3000).
• DEBUG: Enable debug mode (default: false).

💻 Usage Examples

Using Claude Desktop

Add the MCP server to the ~/.config/claude-desktop/mcp-servers.json file:

{
  "mcp-server-rememberizer": {
    "type": "MCP_SERVER",
    "connectionUri": "http://localhost:3000"
  }
}

🔧 Technical Details

Debugging

Since the MCP server runs through standard input and output, debugging can be challenging. For the best debugging experience, we highly recommend using the MCP Inspector.

You can start the inspector with the following command:

npx @modelcontextprotocol/inspector uv --directory /path/to/directory/mcp-servers-rememberizer/src/mcp_server_rememberizer run mcp-server-rememberizer

After starting, the inspector will display a URL in the browser for you to start debugging.

📄 License

This MCP server is distributed under the MIT License. This means you are free to use, modify, and distribute the software, subject to the terms and conditions of the MIT License. For more details, please refer to the LICENSE file in the project repository.