Erinra is a specialized Model Context Protocol (MCP) server that provides long-term memory and structured context storage for AI agents.

How do I resolve the 'Erinra MCP server not connected' error?

You can resolve this by running 'claude mcp add erinra -- erinra serve -s user' in your terminal and then restarting the Claude Code environment.

When should I use the init-memory skill?

You should use this skill when the automatic SessionStart hook fails to load your context, or if Claude seems to have forgotten your recent decisions and preferences.

What kind of data does this skill retrieve?

It retrieves deduplicated memories, user preferences, recent patterns, and a full taxonomy including projects, types, tags, and relationship statistics.

Erinra Memory Initializer

Name: Erinra Memory Initializer
Author: alphaleonis

byalphaleonis

0•

Productivity & Workflow

Manually synchronizes session context and persistent memories using the Erinra MCP server.

The Erinra Memory Initializer provides a reliable manual fallback for loading session context within Claude Code. It is specifically designed for situations where automatic startup hooks fail to trigger, ensuring that Claude has access to essential user preferences, recent project decisions, and deduplicated memories. By interfacing directly with the Erinra MCP server, this skill retrieves a comprehensive taxonomy of projects, types, and tags, allowing for a more personalized and context-aware coding experience.

Key Features

01Full taxonomy synchronization including projects and tags

02Manual fallback for session context initialization

03Deduplicated memory retrieval based on custom queries

04Integrated connectivity health checks for Erinra MCP

050 GitHub stars

06Configurable content budget and retrieval limits

Use Cases

01Restoring session context after a failed automatic startup hook

02Manually refreshing the AI's understanding of project-specific taxonomy

03Synchronizing user preferences and coding patterns across sessions

Key Features

01Full taxonomy synchronization including projects and tags

02Manual fallback for session context initialization

03Deduplicated memory retrieval based on custom queries

04Integrated connectivity health checks for Erinra MCP

050 GitHub stars

06Configurable content budget and retrieval limits

Use Cases

01Restoring session context after a failed automatic startup hook

02Manually refreshing the AI's understanding of project-specific taxonomy

03Synchronizing user preferences and coding patterns across sessions