Memory Server

Qdrant × Quarkus MCP Memory Server
High-performance semantic memory orchestration for agents and daemons.

Overview

Memory Server provides a powerful semantic memory storage and retrieval system tailored specifically for use with the Model Context Protocol (MCP). It leverages the speed of native Quarkus applications, Kotlin for type-safety, and Qdrant for scalable vector-based semantic search.

Features

• Semantic Vector Memory: Store and recall memories based on semantic similarity rather than exact matches.
• Tag-based Filtering: Easily organize and filter memories using flexible tagging.
• Dual CLI and Agent Interface: Designed for both human operators (memory-cli) and AI agents (memory-proxy).
• MCP Compatible: Uses standard MCP calls for seamless integration with AI agents like Goose.

Quickstart

Local Development

To run Memory Server locally using Docker Compose:

docker-compose up

CLI Usage

To use the command-line client for memory operations:

./memory-cli remember --content "This is a test memory" --tags "qa:true,env:dev"
./memory-cli recall --content "test memory"

Agent (Goose) Integration

Load the memory-server extension via Goose CLI with STDIO proxy:

/extension memory-proxy

Example MCP calls via Goose:

{
  "request": {
    "memory": {
      "content": "Semantic search test"
    },
    "page": 1,
    "pageSize": 10
  }
}

Architecture

memory-cli ↔ memory-proxy ↔ REST (memory-server) ↔ gRPC ↔ Qdrant

• memory-server: REST endpoint, semantic embedding generation, gRPC client to Qdrant.
• memory-cli: User/operator CLI.
• memory-proxy: STDIO proxy enabling MCP tool calls from AI agents like Goose.

Recipes & Usage

• QA Recipes available in recipes/.
• Example configuration and recipes provided for immediate integration with Goose CLI.

Contributing

Contributions are welcome! Please open an issue or pull request on GitHub.

License

This project is licensed under the APACHE 2.0 License – see the LICENSE file for details.