mem-search

Search claude-mem's persistent memory index for relevant prior work across sessions.

1|Updated Apr 27, 2026
One-click install
npx skills add https://github.com/dwaynimay/IoTProject --skill mem-search-dwaynimay
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: mem-search
Source: https://github.com/dwaynimay/IoTProject/tree/main/.agents/skills/mem-search
Command: npx skills add https://github.com/dwaynimay/IoTProject --skill mem-search-dwaynimay

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

It helps you quickly determine whether a question, bug fix, or implementation detail was solved in earlier sessions, so you don’t repeat work and can reuse proven context.

Core Features & Use Cases

  • Cross-session memory search: Locate relevant past records using a query and optional filters like type, date range, and project name.
  • Timeline anchoring: Pull surrounding context around one or more key results to understand what led to the outcome.
  • Safe retrieval workflow: Reduce token usage by searching and filtering first, then fetching full details only for selected IDs.
  • Use Case: If you remember “we fixed something related to authentication last week” but can’t recall the exact approach, you can search for prior authentication observations, review the timeline, and then fetch the full details for only the most relevant entries.

Quick Start

Use the mem-search skill to search for prior work on the topic by asking: “Did we already solve authentication token expiration in a previous session, and what was the fix?”

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 sessions for a specific bugfix or decision?

To avoid repeating work, cross-session memory search queries a persistent index using keywords and optional filters like project name, date range, and type to retrieve relevant prior decisions. This mechanism helps you reuse proven context and skip redundant debugging or implementation steps.

What is the best way to retrieve old feature history without wasting tokens?

The best way to retrieve feature history while optimizing tokens is a search-then-filter workflow. You search and filter the memory index first to establish relevance, then fetch full details only for selected IDs, significantly reducing token usage compared to pulling all records immediately.

How does timeline context help with memory search and retrieval?

Timeline context anchors your search by pulling surrounding records around key results, helping you understand what led to a specific outcome or decision. This provides broader context than a single entry, allowing you to trace the progression of a bugfix or feature implementation across sessions.

Can I filter cross-session memory search results by project and date range?

Yes, you can filter cross-session memory search results using optional constraints for a specified project, date range, and record type. These filters narrow the search scope, ensuring you only retrieve observations and decisions relevant to the specific session you are investigating.

Does cross-session memory retrieval require fetching full details for every match?

No, cross-session memory retrieval uses a safe retrieval workflow that avoids fetching full details before relevance is established. It uses MCP tools to search, apply timeline context, and retrieve observations, fetching full records only for the specific IDs you select.