mem-search

Search persistent cross-session memory databases for past Claude Code work.

11|2|Updated Mar 31, 2026
One-click install
npx skills add https://github.com/ajsai47/holyclaude --skill mem-search-ajsai47
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: mem-search
Source: https://github.com/ajsai47/holyclaude/tree/main/skills/memory/mem-search
Command: npx skills add https://github.com/ajsai47/holyclaude --skill mem-search-ajsai47

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill eliminates the frustration of forgotten work between Claude Code sessions, saving you from re-solving problems or re-researching solutions you've already completed in past projects.

Core Features & Use Cases

  • Token-Efficient 3-Layer Search Workflow: Filter results via a lightweight search index first, pull context around relevant entries, then only fetch full details for exact matches, cutting token usage by 10x.
  • Flexible Filtering Options: Narrow searches by project, entry type (observations, sessions, prompts), observation subtype (bugfix, feature, decision), date range, and more to find exactly what you need fast.
  • Use Case: If you can't remember how you fixed an authentication token expiration bug two weeks ago, use this Skill to quickly locate the relevant past work and pull the full implementation details without redoing the work.

Quick Start

Use the mem-search skill to find all past authentication bug fixes in the current project from the last month.

Frequently Asked Questions about mem-search

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I search past Claude Code sessions for previous bug fixes?

You can search past Claude Code sessions for previous bug fixes by querying a persistent cross-session memory database, filtering by observation subtypes like bugfix, project, and date range to retrieve exact implementation details.

What is the best way to find past project decisions without wasting context tokens?

Finding past project decisions without wasting context tokens involves a 3-layer search workflow that filters a lightweight index first, pulls timeline context, then batch fetches full details only for exact matches.

Can I filter cross-session memory searches by specific project and date range?

Yes, you can filter cross-session memory searches by specific project, entry type, observation subtype, and date range to narrow results and locate exact past work entries from previous sessions.

How does the 3-layer search workflow retrieve past work efficiently?

The 3-layer search workflow retrieves past work efficiently by applying a lightweight index filter first, retrieving timeline context around relevant entries, then batch fetching full details only for matched items.

When do I need to search persistent memory for past Claude Code work?

You need to search persistent memory for past Claude Code work when you cannot remember prior solutions, bug fixes, or project decisions and want to avoid redundant research across sessions.