vault-search

Query Obsidian vault content with semantic search and SQL-style metadata filters.

209|29|Updated Jan 3, 2026

Search Tags:#SQL #semantic search #embeddings #metadata #vault #notes #Dataview

One-click install

npx skills add https://github.com/Roasbeef/obsidian-claude-code --skill vault-search-roasbeef

Or copy as Structured Prompt for Agent▼

Please help me install this Agent Skill.
Skill: vault-search
Source: https://github.com/Roasbeef/obsidian-claude-code/tree/main/skills/vault-search
Command: npx skills add https://github.com/Roasbeef/obsidian-claude-code --skill vault-search-roasbeef

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires sqlite-vec, chromadb, pyyaml, and includes scripts (resource) components.

What problem does it solve?

This Skill provides semantic search and SQL-style querying over the Obsidian vault. It helps you find notes by meaning, discover related content, and query frontmatter metadata without manually sifting through files.

Core Features & Use Cases

Semantic search: Find notes by meaning using vector embeddings, returning the most relevant results even if exact keywords differ.
Dataview-style queries: Run SQL-like queries over frontmatter and metadata to build dashboards, lists, and reports.
Combined workflows: Chain semantic search with metadata filters to surface precise notes (e.g., open tasks in a specific folder).

Quick Start

Use the vault-search skill to perform a basic semantic search and a metadata query:

Basic semantic search: python scripts/search.py --query "portfolio risk" --n-results 5
SQL-style metadata query: python scripts/dataview.py --sql "SELECT path, due FROM notes WHERE status = 'open'"