How does Claude find relevant memories?

Claude uses a summary-first search approach powered by ripgrep to quickly scan frontmatter metadata before reading full memory files.

Is the memory shared with other team members?

By default, the memory files are gitignored. This means they are specific to your local environment unless you manually choose to remove them from .gitignore and commit them to your repository.

Where are the memories stored?

Memories are stored locally in your project under the .claude/skills/agent-memory/memories/ directory, keeping your data under your control.

When does the Agent Memory skill trigger?

It triggers on specific phrases like 'remember this' or 'save this note', and Claude can also use it proactively when it discovers valuable findings worth preserving.

Agent Memory

Name: Agent Memory
Author: morixxfoxdata

bymorixxfoxdata

生产力与工作流

Mainages a persistent, organized knowledge base for Claude that survives across different conversations and coding sessions.

关于

Agent Memory provides a structured system for Claude to store, retrieve, and organize critical project information within your local development environment. By leveraging a dedicated local storage directory, it allows the AI to record architectural decisions, complex research findings, and specific implementation patterns that would otherwise be lost when a session ends. It features a robust search workflow using ripgrep and a metadata-rich categorization system that ensures your technical context is always available for future reference, making it an essential tool for long-term project maintenance and complex problem-solving.

主要功能

Structured frontmatter for efficient summary-based searching
0 GitHub stars
Flexible category-based folder organization for memories
Persistent knowledge storage across multiple AI conversations
Integrated ripgrep (rg) workflow for lightning-fast information retrieval
Proactive recording of research findings and codebase patterns

使用场景

Consolidating architectural decisions and the rationale behind specific code implementations
Maintaining project context and 'next steps' during long-term feature development
Documenting non-obvious bugs or dependency issues for future team members or AI sessions

关于

主要功能

Structured frontmatter for efficient summary-based searching
0 GitHub stars
Flexible category-based folder organization for memories
Persistent knowledge storage across multiple AI conversations
Integrated ripgrep (rg) workflow for lightning-fast information retrieval
Proactive recording of research findings and codebase patterns

使用场景

Consolidating architectural decisions and the rationale behind specific code implementations
Maintaining project context and 'next steps' during long-term feature development
Documenting non-obvious bugs or dependency issues for future team members or AI sessions