pieces

Integrate PiecesOS context into developer tools for memory-powered workflows.

1|Updated Nov 25, 2025

Search Tags:#cli #mcp #obsidian #snippets #ltm #pieces #ide-plugins

Ready to run

npx skills add https://github.com/89jobrien/dotfiles --skill pieces

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: pieces
Source: https://github.com/89jobrien/dotfiles/tree/main/dot-claude/skills/pieces
Command: npx skills add https://github.com/89jobrien/dotfiles --skill pieces

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires curl, claude, uv, pieces-cli, python3, and includes scripts (resource) and references (resource) components.

What problem does it solve?

Pieces enables developers to capture, store, and retrieve contextual memory from tools and conversations directly on your device, reducing context-switching and manual note-taking.

Core Features & Use Cases

Long-Term Memory (LTM) context across CLI, IDE plugins, and Obsidian for persistent, searchable workflows.
MCP Integration: connect Pieces to AI clients to access saved snippets and past decisions during coding and reviews.
Snippet Drive & Timeline: save, enrich, and surface code fragments and notes across surfaces for quick reference.

Quick Start

Install PiecesOS, enable Long-Term Memory, and connect a client (CLI, MCP, or IDE plugin) to start saving and querying snippets.

Frequently Asked Questions about pieces

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

FAQPage Schema

How does on-device long-term memory work for AI coding workflows?▼

On-device long-term memory (LTM) captures and stores contextual snippets and conversations locally, allowing AI clients to retrieve past decisions and code fragments during development tasks without manual note-taking.

How do I save and retrieve code snippets across IDE plugins and CLI?▼

Install PiecesOS and connect your IDE plugins or CLI to save code snippets to a local drive. You can then search and retrieve these enriched fragments directly within your development environment.

Does the Pieces MCP integration work with local AI clients?▼

Yes, the Pieces MCP integration connects on-device memory to local AI clients, enabling them to access saved snippets and past contextual decisions directly during coding and code reviews.

What do I need to enable memory-powered workflows in Obsidian?▼

You need PiecesOS running with Long-Term Memory enabled, along with the Pieces MCP and Obsidian integration installed, to capture, search, and leverage persistent contextual memory across your notes.

Can I use local snippet storage to reduce context-switching during code reviews?▼

Yes, local snippet storage and timeline surfaces save code fragments and notes on-device, allowing you to quickly reference past context and decisions without switching away from your review environment.

What are the limitations of relying on on-device memory for development tasks?▼

On-device memory requires PiecesOS to be actively running with LTM enabled, and functionality depends on having the Pieces MCP, IDE plugins, and CLI tools properly installed and connected to provide end-to-end integration.