MCP Catalogs
Homeshodh-memory screenshot

shodh-memory

by varun29ankuS·212·Score 51

Persistent cognitive memory for AI agents and robots with offline support and 37 MCP tools.

ai-llmknowledge-graphproductivity
28
Forks
11
Open issues
this month
Last commit
2d ago
Indexed

Overview

Shodh-Memory is a sophisticated memory system for AI agents and robots that learns from use, forgets irrelevant data, and strengthens important memories. It features a neuroscience-inspired three-tier memory model (working, session, and long-term) with Hebbian learning and activation decay. The system is completely offline with no external dependencies, storing data locally using embeddings and mathematical algorithms rather than LLM API calls. It provides a unified CLI, TUI dashboard, REST API, and extensive MCP tools for memory management, todos, projects, and reminders.

Try asking AI

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

you:Persistent memory for AI coding assistants like Claude and Cursor
you:Cognitive memory for robots and ROS2 systems
you:Personal knowledge management with automatic recall
you:How does Shodh-Memory differ from other memory systems like mem0?
you:Can I use Shodh-Memory with robots?

When to choose this

Choose Shodh-Memory when you need a persistent memory system for AI agents that works fully offline with minimal dependencies, particularly for robotics applications or when low-latency memory operations are critical.

When NOT to choose this

Avoid this if you need cloud-based memory sync across devices, require integration with OpenAI services, or need a system that already has established enterprise support and documentation.

Tools this server exposes

12 tools extracted from the README
  • remember

    Store a new memory with optional type and tags

  • recall

    Search and retrieve memories based on query

  • add_todo

    Create a new todo item

  • list_todos

    Retrieve all todo items

  • update_todo

    Modify an existing todo item

  • complete_todo

    Mark a todo as completed

  • add_project

    Create a new project

  • list_projects

    Retrieve all projects

  • set_reminder

    Create a new reminder

  • list_reminders

    Retrieve all reminders

  • memory_stats

    Get statistics about memory usage

  • backup_create

    Create a backup of all data

Comparable tools

mem0cogneezep

Installation

Installation

**Unified CLI**

# Download from GitHub Releases
shodh init
shodh server

**Claude Desktop**

{
  "mcpServers": {
    "shodh-memory": {
      "command": "npx",
      "args": ["-y", "@shodh/memory-mcp"]
    }
  }
}

**Python**

pip install shodh-memory

**Rust**

[dependencies]
shodh-memory = "0.1"

FAQ

How does Shodh-Memory differ from other memory systems like mem0?
Shodh-Memory uses algorithmic intelligence with local embeddings and mathematical decay, requiring 0 LLM calls for memory storage. It works fully offline, unlike competitors that rely on OpenAI API and external databases.
Can I use Shodh-Memory with robots?
Yes, it has native support for robotics through Zenoh/ROS2 integration with robotics-specific fields like GPS coordinates, sensor data, and mission context. Robots can remember and recall spatial information without cloud connectivity.

On Hacker News

Recent discussion from the developer community.

Compare shodh-memory with

GitHub →

Last updated · Auto-generated from public README + GitHub signals.