skills.rest
softspark
softsparkCOMMUNITY
Explore all skills by this author →
WebsiteGitHub

mem-search

Search past coding session memories using SQLite FTS5 full-text queries.

161|21|Updated Mar 26, 2026
Search Tags:#search#memory#sqlite#knowledge#session#developer#fts5
One-click install
npx skills add https://github.com/softspark/ai-toolkit --skill mem-search-softspark
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: mem-search
Source: https://github.com/softspark/ai-toolkit/tree/main/app/plugins/memory-pack/skills/mem-search
Command: npx skills add https://github.com/softspark/ai-toolkit --skill mem-search-softspark

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Developers often waste time re-discovering why decisions were made or what occurred during prior coding sessions; mem-search lets you quickly locate relevant observations, decisions, and context so you can resume work without manual digging.

Core Features & Use Cases

  • Natural-language search: Enter plain queries to locate past notes, code observations, and decision rationale across sessions.
  • Local full-text indexing: Uses a local SQLite FTS5 index to return ranked observations and associated session metadata.
  • Progressive disclosure: Presents a concise summary view first and reveals full observation details and related session items on demand.
  • Use Case: Recover the reasoning for a past refactor, find a previous bug fix discussion, or surface snippets and context from an earlier debugging session.

Quick Start

Search memory for recent discussion about database migrations to retrieve relevant observations and session summaries.

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 coding session notes and decision rationale?▼

To search past session notes, you enter plain natural-language queries to retrieve ranked observations, edit history, and decision rationale using a local full-text index. This quickly surfaces relevant context to resume work without manual digging.

What is the best way to find the reasoning behind a previous code refactor?▼

The best way to find refactor reasoning is performing a full-text FTS5 query against a local SQLite database of session memories. This returns ranked observations with associated session metadata to help recover past decision context.

Does mem-search work with a local SQLite database for developer workflows?▼

Yes, mem-search works with a local SQLite database using FTS5 to index and retrieve developer session memories. It is designed for local developer workflows where retrieving prior session notes accelerates debugging and feature work.

Can I use natural language queries to find past debugging session context?▼

Yes, you can use natural-language queries to find past debugging session context. The search performs full-text queries against a local SQLite database and returns ranked observations with session metadata while supporting query escaping.

How are full observation details revealed during a session memory search?▼

Full observation details are revealed through progressive disclosure, which presents a concise summary view first and then reveals full observation details and related session items on demand. This keeps initial search results readable.

Related Skills

Explore More Software Engineering Skills→
StevenWXYStevenWXY

sqlite-local-storage

Store file metadata with FTS5 full-text and sqlite-vec vector search.

Community
Advanced
cheng80cheng80
1

context-restore

Restore the latest saved context across branches and Conductor handoffs.

Community
Basic
MSApps-MobileMSApps-Mobile
8

cowork-mem

Store and recall persistent session memory in SQLite with FTS5 and TF-IDF search.

Official
Advanced
zli5460zli5460
2

qmd

Search local knowledge bases with hybrid keyword and semantic queries.

Community
Intermediate