hoard-memory

Search, read, and write notes in Ember's persona memory vault via storybook-daemon MCP.

Updated Mar 31, 2026
One-click install
npx skills add https://github.com/dotBeeps/hoard --skill hoard-memory
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: hoard-memory
Source: https://github.com/dotBeeps/hoard/tree/main/cc-plugin/skills/memory
Command: npx skills add https://github.com/dotBeeps/hoard --skill hoard-memory

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Provide agents with a searchable, persistent memory store for Ember by exposing the storybook-daemon memory vault so observations, decisions, and session context can be recorded and retrieved across interactions.

Core Features & Use Cases

  • Search stored notes by query and limit to quickly surface prior observations and decisions.
  • Read full note content including metadata to review context, tags, and timestamps.
  • Write structured notes with a slug-style title, kind, content, and tags to persist insights and choices.
  • Use cases include preserving architectural decisions, capturing surprising behavior during debugging, and resurfacing relevant prior context at the start of a session.

Quick Start

Invoke the hoard-memory skill to write a note by calling mcp__storybook-ember__memory_write with a slug title, kind, content, and tags.

Frequently Asked Questions about hoard-memory

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

FAQPage Schema
How do I persist agent session context and decisions across interactions?

You persist agent context by writing structured notes with a slug title, kind, content, and tags to a memory vault. This records observations and decisions for retrieval in later interactive or automated sessions.

What is a memory vault for AI agents and how does it work?

An agent memory vault is a searchable, persistent store exposing search, read, and write operations via MCP. It lets agents record discoveries and retrieve prior session observations by querying stored notes.

How do I search and read previous notes in an agent memory vault?

Search stored notes by passing a query string and limit to surface relevant observations. Reading a note retrieves its full content along with metadata like context, tags, and timestamps for review.

Can I use MCP to record architectural decisions during automated debugging sessions?

Yes, you can use an MCP memory vault to record architectural decisions during automated debugging. You write notes with a slug-style title, kind, content, and tags to capture surprising behavior and preserve insights.

What parameters do I need to write a note to a memory vault via MCP?

Writing a note via MCP requires a slug-style title, a kind classification, the content body, and tags. These parameters structure the observation for searchable retrieval within the memory vault.

When should I use a persistent memory store instead of standard session context?

Use a persistent memory store when you need to surface prior context at the start of a new session or preserve insights across automated interactions. It is necessary when discoveries must outlast a single session.