memory-mcp

SKILL.md

Memory MCP Server

Interacts with and troubleshoot the Memory Model Context Protocol (MCP) server for the self-learning memory system.

Quick Reference

When to Use

  • Starting or configuring the memory-mcp server
  • Using MCP tools for memory retrieval and pattern analysis
  • Validating the MCP server implementation
  • Debugging MCP server issues
  • Testing MCP tools using the MCP inspector
  • Understanding MCP configuration and environment variables

MCP Server Overview

The memory-mcp server exposes episodic memory functionality through the Model Context Protocol:

  • Query past experiences and learned patterns
  • Analyze successful strategies from historical episodes
  • Execute code in a secure sandbox environment
  • Perform advanced statistical and predictive analysis
  • Monitor server health and metrics

Location: ./target/release/memory-mcp-server Configuration: .mcp.json Transport: stdio (Standard Input/Output)

Available MCP Tools

Tool Purpose
query_memory Query episodic memory for relevant past experiences
analyze_patterns Analyze patterns from past episodes
advanced_pattern_analysis Statistical analysis, predictive modeling
execute_agent_code Execute TypeScript/JavaScript in sandbox
health_check Check server health status
get_metrics Get comprehensive monitoring metrics

See tools.md for detailed tool documentation and best-practices.md for usage guidelines.

Starting the Server

# Build
cargo build --release --bin memory-mcp-server

# Run directly
export TURSO_DATABASE_URL="file:./data/memory.db"
./target/release/memory-mcp-server

# Run via MCP Inspector for testing
npx -y @modelcontextprotocol/inspector ./target/release/memory-mcp-server

See configuration.md for full environment setup and validation.md for MCP Inspector workflow.

Weekly Installs
20
GitHub Stars
5
First Seen
Feb 21, 2026
Installed on
opencode20
claude-code20
github-copilot20
codex20
amp20
kimi-cli20