Memory

Organize long-term information into categorized markdown files under ~/memory/.

Updated May 4, 2026
One-click install
npx skills add https://github.com/234194027-cpu/xianclaw --skill memory-234194027-cpu
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: Memory
Source: https://github.com/234194027-cpu/xianclaw/tree/main/resources/skills/memory
Command: npx skills add https://github.com/234194027-cpu/xianclaw --skill memory-234194027-cpu

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) and assets (resource) components.

What problem does it solve?

Keeps an unlimited, well-organized long-term memory store for your agent so important information doesn’t get lost or become unfindable.

Core Features & Use Cases

  • Infinite categorized storage: Create user-defined categories under ~/memory/ to match how you think and work.
  • Instant navigability with indices: Maintain INDEX.md at the root and inside each category so you can locate information quickly.
  • Optional one-way sync: Copy selected items from the agent’s built-in memory into organized files without overwriting built-in sources.

Quick Start

Ask the agent to set up your personal memory categories and then store your first item immediately into ~/memory/ using the Memory skill.

Frequently Asked Questions about Memory

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

FAQPage Schema
How do I store long-term project histories and contact profiles for my AI agent?

To store long-term project histories and contact profiles, you can use a categorized knowledge management system that organizes durable notes into an infinite memory structure. This approach maintains durable records in structured markdown entries under user-defined categories.

What is the best way to organize an unlimited knowledge base for an agent without overwriting built-in memory?

The best way to organize an unlimited knowledge base without overwriting built-in memory is using a one-way sync approach. This copies selected items from the agent's built-in memory into organized files within a dedicated root folder, preserving original sources.

How do I set up a categorized memory folder structure for long-term storage?

To set up categorized memory for long-term storage, create a root ~/memory/ folder containing config.md and INDEX.md files. You then define per-category folders with their own INDEX.md files to maintain instant navigability for per-item markdown entries.

Does this long-term storage approach require syncing to modify the agent's built-in memory?

No, this long-term storage approach does not modify built-in memory. The optional one-way sync feature only copies selected items from built-in memory into organized files via a designated sync folder, ensuring original sources remain untouched.

Can I create custom categories to match how I manage knowledge for my projects?

Yes, you can create user-defined categories under the root memory folder to match how you think and work. This allows you to structure decision records, knowledge bases, and other structured notes that grow over time according to your specific workflow.

How does indexing help me find information in a large categorized memory store?

Indexing helps you find information quickly by maintaining INDEX.md files at the root level and inside each category. This provides instant navigability across an infinite categorized memory store, allowing you to locate specific markdown entries without searching blindly.