Skip to content
Conduit

skill

obsidian-claude-tools

A collection of Claude Code plugins for enhancing Obsidian workflows.

tsuyoshimishina/obsidian-claude-toolscommunitylistedunreviewed

Set up

Skills are installed by placing their folder where your agent loads skills.

The download comes from github.com/tsuyoshimishina/obsidian-claude-tools itself, not from our server, so what you get is exactly what its author published.

Install instructions come from the README

Skills do not use an MCP config. Follow the steps in the README below.

README

synced Oct 23, 2026

Obsidian Claude Tools

A collection of Claude Code plugins for enhancing Obsidian workflows.

Overview

This project provides tools that integrate Obsidian with Claude Code, enabling AI-assisted note management and automation. Each tool is implemented as a Claude Code plugin with slash commands, agents, and skills.

Available Tools

Auto-Tagger (/tag-notes)

Automatically assigns tags to Obsidian notes based on their content.

Features:

  • Assigns tags from predefined categories that clearly match note content
  • Fast batch processing with parallel execution (up to 10 concurrent)
  • Dry-run mode for simulation
  • Preserves existing tags while adding new ones

Usage:

/tag-notes                              # Process entire vault
/tag-notes inbox                        # Process inbox folder
/tag-notes notes/projects               # Process specific subfolder
/tag-notes . --dry-run                  # Simulate on entire vault
/tag-notes --criteria custom.md         # Use custom criteria file
/tag-notes inbox --criteria papers.md   # Combine folder + criteria

Tag Criteria:

Model Configuration:

The auto-tagger agent uses opus model by default for highest accuracy. To change the model, edit agents/obsidian-tagger.md:

model: opus  # Options: opus, sonnet, haiku
  • opus - Default, highest accuracy (recommended)
  • sonnet - Good balance of cost and accuracy (95% match with opus in testing)
  • haiku - Fastest and cheapest, may reduce tagging accuracy

Japanese Formatter (/format-japanese-note)

Applies consistent Japanese formatting rules to Obsidian notes.

Features:

  • Converts half-width colons to full-width (:→:)
  • Inserts spaces at half-width/full-width boundaries (VGGTは→VGGT は)
  • Converts half-width parentheses to full-width ((説明)→(説明))
  • Preserves URLs, code blocks, and Markdown syntax
  • Dry-run mode for preview

Usage:

/format-japanese-note inbox/note.md              # Format single file
/format-japanese-note inbox                      # Format all files in folder
/format-japanese-note inbox/note.md --dry-run   # Preview changes

Formatting Rules:

Mind Map Extractor (/extract-mindmap)

Extracts hierarchical structure from NotebookLM Mind Map images and converts to Mermaid format.

Features:

  • Analyzes NotebookLM Mind Map images
  • Outputs Mermaid mindmap syntax
  • Preserves original node text and hierarchy
  • Supports Japanese and English content

Usage:

/extract-mindmap ./downloads/NotebookLM_Mind_Map.png
/extract-mindmap ~/Desktop/mindmap.png --output inbox/project.mermaid.md

Mind Map Memofy (/memofy-mindmap)

Transforms Mermaid mind maps into detailed Markdown notes by enriching each leaf node with web search results.

Features:

  • Parses Mermaid mindmap structure
  • Web search enrichment for each leaf node
  • Parallel processing (up to 10 concurrent)
  • Auto-generates context description
  • Respects copyright (no verbatim copying)

Usage:

/memofy-mindmap inbox/project.mermaid.md
/memofy-mindmap inbox/project.mermaid.md --output notes/project-overview.md
/memofy-mindmap inbox/project.mermaid.md --context "Project X overview"
/memofy-mindmap inbox/project.mermaid.md --max-chars 300

Typical Workflow:

  1. Generate Mind Map in NotebookLM
  2. /extract-mindmap to convert image → Mermaid
  3. (Optional) Edit Mermaid in Obsidian
  4. /memofy-mindmap to enrich → Markdown note

Background:

I was impressed by how NotebookLM's Mind Map feature captures the structure of source materials in a brilliant top-down manner. I realized that enriching these mind maps with additional detail could produce excellent reference documents. This led me to implement these commands. So far, I have been satisfied with the results!

Requirements

  • Claude Code CLI
  • Obsidian MCP Tools (obsidian-mcp-tools)
  • Bash (Linux/macOS native, Windows via Git Bash)

Installation

Run from Git Bash:

./deploy.sh
./deploy.sh --uninstall  # to uninstall

Project Structure

obsidian-claude-tools/
├── .beads/                       # Issue tracking (beads)
├── .claude-plugin/
│   └── plugin.json               # Plugin metadata
├── agents/
│   ├── obsidian-tagger.md        # Auto-tagger agent
│   └── leaf-describer.md         # Mind map leaf enrichment agent
├── commands/
│   ├── tag-notes.md              # /tag-notes command
│   ├── format-japanese-note.md   # /format-japanese-note command
│   ├── extract-mindmap.md        # /extract-mindmap command
│   └── memofy-mindmap.md         # /memofy-mindmap command
├── skills/
│   └── tag-criteria/
│       ├── SKILL.md              # Judgment rules
│       └── tags/
│           ├── _all.md           # Default tag definitions
│           └── *.md              # Custom criteria files
├── deploy.sh                     # Deployment script
└── README.md

Note on .beads/: This project uses beads for issue tracking. If you are not using beads, you can safely ignore the .beads directory.

License

MIT

↑↓ navigate openesc close