memcord

Overview

Memcord is a comprehensive MCP server designed for organizing and managing AI chat history with strong privacy controls. It provides tools for saving, summarizing, searching, and organizing conversations across multiple memory slots. The server supports various summarization backends including extractive methods (NLTK, Sumy) and abstractive models (transformers, semantic), allowing users to balance speed and quality based on their needs.

Try asking AI

After installing, here are 5 things you can ask your AI assistant:

• memcord_namememcord_name(slot_name: string)

  Create or select a memory slot for storing conversations

  Try:Create a memory slot called project_meetingTry:Switch to my_project_name slot
• memcord_savememcord_save(text: string)

  Save exact text to the current memory slot

  Try:Save our discussion about API designTry:Record this conversation summary
• memcord_save_progressmemcord_save_progress()

  Save a compressed summary of current conversation

  Try:Save a summary of our discussionTry:Compress and save this session
• memcord_readmemcord_read(slot_name: string = null)

  Read contents from the current memory slot

  Try:Show me all saved messagesTry:Read the project_meeting memory
• memcord_select_entrymemcord_select_entry(timestamp: string)

  Navigate to a specific point in the conversation timeline

  Try:Jump to the entry from 2 hours agoTry:Go back to yesterday's discussion
• memcord_listmemcord_list()

  List all available memory slots

  Try:Show all my memory slotsTry:List all saved conversations
• memcord_searchmemcord_search(query: string)

  Full-text search across all memory contents

  Try:Search for API design discussionsTry:Find all mentions of authentication
• memcord_querymemcord_query(question: string)

  Ask natural language questions about saved conversations

  Try:What did we decide about API design?Try:Tell me about our authentication strategy
• memcord_initmemcord_init(path: string, slot_name: string)

  Bind a memory slot to the current directory

  Try:Initialize a slot called my-project for this directoryTry:Set up project memory here
• memcord_zeromemcord_zero()

  Enable privacy mode to prevent saving conversations

  Try:Turn off saving for this sessionTry:Enable privacy mode
• memcord_configurememcord_configure(action: string, key: string = null, value: string = null)

  Configure settings for memory slots

  Try:Change summarizer to BARTTry:Set sumy_algorithm to lsa

Comparable tools

Installation

curl -fsSL https://github.com/ukkit/memcord/raw/main/install.sh | bash

irm https://github.com/ukkit/memcord/raw/main/install.ps1 | iex

{
  "mcpServers": {
    "memcord": {
      "command": "python",
      "args": ["-m", "memcord.server"]
    }
  }
}

FAQ

Where is my data stored?

All data is stored locally on your device in your home directory under ~/.memcord. Nothing is sent to the cloud.

How many summarization backends are available?

Memcord supports four backends: NLTK (fast, extractive), Sumy (fast, graph-based extractive), Semantic (medium speed, best extractive), and Transformers (slow, best abstractive quality).

Compare memcord with