semantic-search

Search markdown notes by semantic meaning using embeddings and cached vector indexes.

12|2|Updated Jun 8, 2026

Search Tags:#semantic search #embeddings #knowledge base #duplicate detection #note retrieval #vector index

Ready to run

npx skills add https://github.com/Loringtonian/second-brain-template --skill semantic-search-loringtonian

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: semantic-search
Source: https://github.com/Loringtonian/second-brain-template/tree/main/.claude/skills/semantic-search
Command: npx skills add https://github.com/Loringtonian/second-brain-template --skill semantic-search-loringtonian

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill helps you find the right notes, ideas, and files when exact keywords are not enough, making it easier to recall related content from a second brain.

Core Features & Use Cases

Meaning-based retrieval: Search for concepts instead of matching only literal words.
Duplicate and overlap checks: Verify whether a new idea already exists or partially overlaps with stored notes.
Similarity discovery: Find files that are semantically close to an existing note and surface unexpected connections.
Use case: A user asks for ideas about coordination, and the Skill returns relevant notes even if they use different wording than the query.

Quick Start

Ask the skill to search for a topic in your second brain, then review the ranked results and optionally narrow them by project or category.

Frequently Asked Questions about semantic-search

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

FAQPage Schema

How do I find related notes by meaning instead of exact keywords?▼

Semantic search uses local embedding models to find related notes by conceptual meaning. It retrieves markdown files based on semantic similarity rather than literal keyword matching, enabling meaning-based retrieval across your knowledge base.

Can I check if a new idea is a duplicate of existing notes?▼

Yes, semantic search detects duplicate ideas and partial overlaps by comparing rephrased concepts. It checks whether a new idea already exists in your second brain by finding semantically similar notes, even if the wording differs.

Do I need a local embedding model to search my markdown files?▼

Yes, semantic search requires a local embedding model and a cached vector index to operate. These components process your markdown files and generate the semantic embeddings needed for script-driven search and similarity discovery.

How does semantic similarity discovery work across a second brain?▼

Semantic similarity discovery finds files that are conceptually close to an existing note, surfacing unexpected connections. It ranks related markdown files using semantic embeddings and allows narrowing results by project or category filters.

What is the best way to search for concepts in a knowledge base when exact terms fail?▼

Semantic search provides meaning-based retrieval to find notes when exact keywords are not enough. It applies a cached vector index to broad recall queries, returning ranked results that match the underlying concept of your search.

Can I filter note retrieval results by project or category?▼

Yes, semantic search supports optional project or category filters to narrow note retrieval results. After performing a script-driven search operation, you can refine the ranked semantic matches to focus on specific sections of your knowledge base.