Claude Obsidian Skills

Claude Obsidian Skills

Reusable AI agent skills for Obsidian and Markdown workflows with Claude Code, Gemini CLI, and other AI assistants.

Overview

Claude Obsidian Skills is a collection of AI agent skills designed to automate personal knowledge management (PKM) workflows. These skills work with Claude Code, Gemini CLI, and other AI assistants that support skill-based extensions.

Available Skills

PKM Management Skills

Skill	Description
ai4pkm-helper	AI4PKM helper for onboarding guidance, quick help, and seamless handoff to DDA for daily use
gobi-onboarding	Gobi Desktop 3.0 voice-based onboarding guide with 4-step flow

Obsidian Skills

Skill	Description
obsidian-canvas	Create and manage Obsidian Canvas files with automatic layout generation for knowledge maps and visual summaries
obsidian-links	Format, validate, and fix wiki links with proper filename, section, and folder conventions
obsidian-yaml-frontmatter	Standardize YAML frontmatter properties with consistent formatting and naming
obsidian-markdown-structure	Validate and enforce markdown document structure, heading hierarchy, and organization
obsidian-mermaid	Create Obsidian-compatible Mermaid diagrams avoiding common errors (markdown in labels, wide layouts)

Markdown Skills

Skill	Description
markdown-slides	Create Deckset/Marp compatible presentation slides from markdown content
interactive-writing-assistant	AI-powered writing companion with outline-prose co-evolution and PKM integration

Video Skills

Skill	Description
markdown-video	Convert markdown slides to MP4 video with TTS narration (with or without Deckset)
video-cleaning	Remove silent pauses from videos using AI-powered transcription and FFmpeg
create-vlog-script	Create narration-first personal vlog scripts and slide-based recording guides from rough notes and media
remotion-vlog	Turn camera footage into a concise, review-first Remotion vlog with maps, captions, music, and verified delivery
remotion-narrated-video-essay	Edit narration-led Remotion essays with selective B-roll, face-safe overlays, and still-frame QA
voice-clone-skill	Prepare, generate, and verify local Qwen3-TTS cloned-voice narration; includes the repository owner's authorized jin reference profile

Image Skills

Skill	Description
gemini-image-skill	Generate AI images using Google Gemini API with multiple models and aspect ratios

Installation

For Claude Code

1. Clone this repository or copy the skill folders:

git clone https://github.com/jykim/claude-obsidian-skills.git

2. Copy skills to your Claude Code skills directory:

cp -r claude-obsidian-skills/*-* ~/.claude/skills/

3. Claude Code will automatically detect and use the skills.

For Other AI Assistants

Copy the SKILL.md file from the desired skill folder to your assistant's skill directory. Each skill is self-contained and documented within its SKILL.md file.

Advanced: Git Submodule Integration

For vault maintainers who want to keep skills in sync with this repository while avoiding duplication:

1. Add as a submodule to your vault:

git submodule add https://github.com/jykim/claude-obsidian-skills.git path/to/skills/public-skills

2. Create symlinks for backward compatibility:

cd path/to/skills
ln -s public-skills/obsidian-canvas obsidian-canvas
ln -s public-skills/obsidian-links obsidian-links
ln -s public-skills/obsidian-yaml-frontmatter obsidian-yaml-frontmatter
ln -s public-skills/obsidian-markdown-structure obsidian-markdown-structure
ln -s public-skills/obsidian-mermaid obsidian-mermaid
ln -s public-skills/markdown-slides markdown-slides
ln -s public-skills/markdown-video markdown-video
ln -s public-skills/interactive-writing-assistant interactive-writing-assistant
ln -s public-skills/video-cleaning video-cleaning

3. Update skills when new versions are released:

git submodule update --remote

Benefits:

• Single source of truth - edit in one place
• Version control - pin to specific commits if needed
• Easy updates via git submodule commands

Note: When cloning a vault with submodules, use --recursive:

git clone --recursive your-vault-repo

Requirements

Some skills require external dependencies. Install them before using:

Skill	Requirements
markdown-video	FFmpeg, OpenAI API key, Python 3.7+, Pillow
video-cleaning	FFmpeg, OpenAI API key, Python 3.7+
create-vlog-script	FFmpeg and a supplied media folder
remotion-vlog	Node.js, Remotion, FFmpeg
remotion-narrated-video-essay	Node.js, Remotion, FFmpeg
voice-clone-skill	Apple Silicon, Python 3.11, FFmpeg, Qwen3-TTS/MLX runtime and model
gemini-image-skill	Gemini API key, Python 3.7+, google-genai, Pillow
markdown-slides	None (markdown only)
obsidian-*	None (file operations only)
interactive-writing-assistant	None

Common Installation Commands

# FFmpeg (macOS)
brew install ffmpeg

# Python dependencies
pip install Pillow openai google-genai

# Set API keys
export OPENAI_API_KEY="sk-..."
export GEMINI_API_KEY="your-gemini-api-key"

Quick Start

Validate Wiki Links

Use the obsidian-links skill to check and fix broken wiki links in my vault

Standardize Frontmatter

Use obsidian-yaml-frontmatter to ensure consistent properties across my notes

Create Mermaid Diagrams

Use obsidian-mermaid to create a flowchart for this process

Create Visual Canvas

Use obsidian-canvas to create a knowledge map for this topic

Create Presentation

Use markdown-slides to convert this document into a Deckset presentation

Generate Video

Use markdown-video to create a narrated video from my slides

Clean Video

Use video-cleaning to remove silent pauses from this video

Create a Remotion Vlog

Use remotion-vlog to turn this camera folder into a concise vlog and stop for preview approval before the final render

Generate Cloned-Voice Narration

Use voice-clone-skill to generate and verify this Korean narration with my authorized local voice profile

Voice Profile Notice

The bundled jin voice profile is an authorized reference sample from the repository owner. It is included for transparent experimentation with the voice-cloning skill. Do not use it to deceive, impersonate the speaker, imply endorsement, or bypass consent and applicable law. Generated audio should be disclosed as synthetic when context could otherwise mislead listeners.

Skill Structure

Each skill follows a consistent structure:

skill-name/
├── SKILL.md          # Main skill definition and documentation
├── reference/        # Reference materials (optional)
└── examples/         # Usage examples (optional)

The SKILL.md file contains:

• Skill metadata (name, description, allowed tools)
• When to use the skill
• Core rules and guidelines
• Step-by-step workflows
• Quality checklists

Contributing

We welcome contributions! Here's how you can help:

1. Report bugs - Open an issue describing the problem
2. Suggest improvements - Share ideas for enhancing existing skills
3. Add new skills - Create skills for other PKM workflows
4. Improve documentation - Help clarify usage and examples

Creating a New Skill

1. Create a folder in the repo root with your skill name
2. Add a SKILL.md with:
   • Frontmatter (name, description, allowed-tools, license)
   • Clear "When to Use" section
   • Step-by-step process documentation
   • Quality checklist
3. Test the skill with Claude Code or your preferred assistant
4. Submit a pull request

Related Projects

• AI4PKM - Full PKM automation framework
• AI4PKM Docs - Documentation and guides
• Discord Community - Join the discussion

AI4PKM Framework

These skills are part of the AI4PKM framework for AI-powered personal knowledge management. For the complete automation system including workflows, templates, and integrations, check out the main project.

License

MIT License - see LICENSE for details.

---

Made with AI assistance for AI assistance.