What is the purpose of the Remember skill?

It allows you to save information—like preferences, patterns, or bug fixes—into a persistent database so Claude can recall that context in future sessions.

How does it handle similar or duplicate memories?

The skill automatically checks for the top-3 most similar existing memories. If a match is found, it can update, merge, or skip the new entry to keep your database clean.

Is it safe to store API keys with this skill?

No. You should never store secrets, credentials, or sensitive data in the memory system.

Where is the memory data actually stored?

Data is stored in your Erinra MCP server instance. You must have the erinra server connected to Claude Code for this skill to function.

Should I store project-specific documentation here?

It is generally better to store project-specific patterns in a local CLAUDE.md file. Use the Remember skill for cross-project knowledge and personal preferences.

Decaf Memory Remember

Name: Decaf Memory Remember
Author: alphaleonis

byalphaleonis

0•

생산성 및 워크플로

Persists user preferences, coding patterns, and architectural decisions to a long-term knowledge base using the Erinra MCP server.

The Remember skill provides Claude with a long-term memory system by integrating with the Erinra MCP server. It allows users to capture important context—such as personal coding styles, complex bug fixes, and high-level architectural decisions—that would otherwise be lost between sessions. The skill intelligently classifies information, checks for duplicates, and links related concepts together, creating a structured knowledge base that makes Claude more helpful and personalized as you use it. It includes built-in safeguards to prevent the storage of sensitive data or project-specific notes that belong in local documentation.

주요 기능

01Project-aware tagging and organization for easy retrieval

02Structured safeguards against storing secrets or transient data

03Intelligent deduplication and merging of similar existing memories

040 GitHub stars

05Automatic classification into types like preference, pattern, decision, and bug-fix

06Relational linking to connect new memories to existing context

사용 사례

01Documenting obscure bug fixes and their solutions to prevent recurring issues

02Recording global development preferences like indentation styles or naming conventions

03Saving high-level architectural choices that span multiple projects and repositories

주요 기능

01Project-aware tagging and organization for easy retrieval

02Structured safeguards against storing secrets or transient data

03Intelligent deduplication and merging of similar existing memories

040 GitHub stars

05Automatic classification into types like preference, pattern, decision, and bug-fix

06Relational linking to connect new memories to existing context

사용 사례

01Documenting obscure bug fixes and their solutions to prevent recurring issues

02Recording global development preferences like indentation styles or naming conventions

03Saving high-level architectural choices that span multiple projects and repositories