Advertise
MCP.so
Sign In

MCP Synaptic

@jvanmelckebeke

Repository

Basic information

Category

Memory & Knowledge

Transports

stdio

Author

jvanmelckebeke

About MCP Synaptic

Memory-enhanced MCP server with local RAG database and expiring memory capabilities

#mcp-synaptic#memory-enhanced#rag-database

MCP Synaptic config

Add this server to your MCP-compatible client using the configuration below.

{
  "mcpServers": {
    "mcp-synaptic": {
      "command": "uv",
      "args": [
        "sync"
      ]
    }
  }
}

MCP Synaptic tools

Tool list not available yet

This listing does not currently have a tool list to display. This does not mean the server has no tools; check its documentation for supported capabilities.

We auto-extract tools from the README. The maintainer can list them under a ## Tools heading to populate this section.

Overview

What is MCP Synaptic?

MCP Synaptic is a memory-enhanced MCP (Model Context Protocol) server that provides a local RAG (Retrieval-Augmented Generation) database and expiring memory capabilities. It is designed for developers who need to store temporary memories with configurable TTL, manage documents via vector search, and stream real-time events—all within the MCP ecosystem.

How to use MCP Synaptic?

Install Python 3.11+ and the UV package manager. Run uv sync to install dependencies, then uv run mcp-synaptic init to initialize the project and uv run mcp-synaptic server to start the server on http://localhost:8000. For containerized deployment, use docker-compose up --build. Configuration is managed through environment variables (see .env.example).

Key features of MCP Synaptic

  • Expiring memories with configurable TTL and automatic cleanup
  • Support for ephemeral, short-term, long-term, and permanent memory types
  • Optional Redis backend for distributed memory storage
  • ChromaDB-based vector database with semantic search
  • API-based or local embedding models (sentence-transformers)
  • Full MCP protocol, SSE, and WebSocket real-time communication
  • Docker-ready with multi-service orchestration

Use cases of MCP Synaptic

  • Store user preferences or session data that automatically expire after a set time
  • Build AI assistants that can retrieve relevant documents via semantic search
  • Develop multi-agent systems that share short-term context through memory
  • Deploy memory-backed MCP servers in containerized environments

Frequently asked questions

1

What embedding options does MCP Synaptic support?

MCP Synaptic supports API-based embeddings (lightweight, works with OpenAI-compatible endpoints) and local embeddings using PyTorch and sentence-transformers (heavier but no external dependency).

2

What is the default TTL for memories and how can I change it?

The default memory TTL is 3600 seconds (1 hour). You can change it via the `DEFAULT_MEMORY_TTL_SECONDS` environment variable.

3

Can I use MCP Synaptic without Docker?

Yes. You can run MCP Synaptic directly with Python 3.11+ and UV. Docker is optional and only needed for containerized deployments.

4

What happens when a memory expires?

Expired memories are automatically removed by a background cleanup process that runs at a configurable interval (default 300 seconds). Events are emitted via SSE when a memory expires.

5

Does MCP Synaptic require Redis?

No. Redis is optional and only needed if you want a distributed memory store. By default, MCP Synaptic uses SQLite and local ChromaDB storage.

Comments

More Memory & Knowledge MCP servers

UmmahAPI

Yousuf Kazmi

Islamic source data for assistants: Quran, hadith, tafsir, duas, prayer times and qibla. Answers come from real sources with citations instead of recalled text, so verses and hadith are not hallucinated. 36,000+ hadiths

Scento Perfume API (MCP server): free fragrance API, no key

Scento

Scento Perfume API is a free fragrance API and perfume notes database for developers and AI agents. Search 1,500+ perfumes and 400+ ingredients, compare scents, find similar perfumes, map scent families and build occasio

Stele

Stele-Dev

Shared memory for AI coding agents. Claude Code, Cursor and Codex read your project's decisions, lessons and tasks before they act.

M1K3

Round-Tower

Local AI agent for your Mac with a built-in MCP server: 18 tools for your notes and PDFs, a temporal memory graph, an on-device LLM and a neural voice. No account, no telemetry. No data centers required.

7

Neither MCP

stonianua

Project context your AI can query through MCP — with source evidence.

Advertise