qmd

Search local Canvas Notebook workspace files using BM25 with vector fallback.

10|2|Updated Jan 12, 2026

Search Tags:#documents #search #workspace #notes #vector #bm25 #qmd

Ready to run

npx skills add https://github.com/canvascoding/canvas-notebook --skill qmd-canvascoding

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: qmd
Source: https://github.com/canvascoding/canvas-notebook/tree/main/seed_skills/qmd
Command: npx skills add https://github.com/canvascoding/canvas-notebook --skill qmd-canvascoding

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Local workspace search across Canvas Notebook to quickly locate notes, files, and documents, including derived content from scanned or converted sources.

Core Features & Use Cases

Fast, text-based search across workspace data using BM25 by default.
Supports derived content extraction to surface notes and documents across projects.
Use Case: Find a related note or a specific document across team folders.

Quick Start

Ask the AI to search the workspace for a term and surface matching notes and documents.

Frequently Asked Questions about qmd

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

FAQPage Schema

How do I search for notes and documents across my local workspace?▼

You can search for notes and documents across a local workspace by applying BM25 text-based search by default, with vector search fallback as needed to surface matching files across multiple projects and file types.

What is the best way to find related notes across different team folders?▼

Finding related notes across team folders requires a workspace search tool that supports derived content extraction, allowing you to quickly locate specific documents and surface text from scanned or converted sources.

Does workspace search require specific runtime environments to function?▼

Yes, workspace search requires a runtime environment satisfying Bun >= 1.0.0 and SQLite3 to execute local text searches and manage the underlying workspace data effectively.

Can I search for derived content extracted from scanned documents?▼

Yes, searching for derived content extracted from scanned or converted documents is supported, allowing you to surface and locate files that originate from non-text sources across your projects.

Why use BM25 search instead of vector search for locating workspace files?▼

BM25 is used by default for locating workspace files because it provides fast, safe text-based search results, while vector search serves as an automatic fallback when query complexity requires semantic matching.

Does qmd support searching both workspace-text and workspace-derived sources?▼

Yes, qmd supports searching both workspace-text and workspace-derived sources, ensuring comprehensive coverage when locating notes and files across your local Canvas Notebook environment.