memory

Store and retrieve user context across sessions in a local qmd-backed memory store.

1|1|Updated Aug 3, 2020
One-click install
npx skills add https://github.com/lolwierd/dotfiles --skill memory-lolwierd
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: memory
Source: https://github.com/lolwierd/dotfiles/tree/main/ai/dot-agents/skills/memory
Command: npx skills add https://github.com/lolwierd/dotfiles --skill memory-lolwierd

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Persistent, searchable memory that survives across sessions, enabling the assistant to recall user preferences, past decisions, and contextual details without repeated prompts.

Core Features & Use Cases

  • Persistent, searchable memory store at ~/.pi/agent/memory/
  • Organizes memories by topic/category and supports keyword and semantic search
  • Automatically saves learnings like preferences and decisions for future recall
  • Ideal for onboarding new tasks, project handoffs, and long-running conversations

Quick Start

Create a memory entry describing a user preference under ~/.pi/agent/memory/ and ensure it is indexed for fast retrieval.

Frequently Asked Questions about memory

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

FAQPage Schema
How do I persist context and recall user preferences across chat sessions?

To persist context across sessions, you can use a local, searchable memory store that automatically saves user preferences and prior decisions. This allows an assistant to recall past details without repeated prompts.

How does local memory persistence work for long-running projects?

Local memory persistence works by storing contextual details and decisions in a local, searchable directory indexed for fast retrieval. It organizes memories by topic to support keyword and semantic search during project handoffs.

Do I need a specific local directory setup to store agent memory?

Yes, you need a local searchable store located at ~/.pi/agent/memory/. You create memory entries describing preferences or decisions in this directory and ensure they are indexed for fast retrieval.

What is the best way to avoid repeating decisions in design chats?

The best way to avoid repeating decisions is to use a persistent memory store that automatically saves learnings. It enables keyword and semantic search to recall past decisions and user preferences for future guidance.

Can I use keyword and semantic search to retrieve past context?

Yes, the memory store supports both keyword and semantic search to retrieve past context. It organizes memories by topic and category, allowing you to find specific preferences and decisions quickly.