claude-knowledge-catalyst 0.10.0

AI-powered knowledge management with YAKE keyword extraction for Claude Code development

Claude Knowledge Catalyst (CKC) v0.10.0

Claude Code ⇄ Obsidian Seamless Integration System

Automatically synchronize insights from Claude Code development processes with Obsidian vaults for structured knowledge management. Automated analysis reduces manual classification overhead.

📋 Japanese Version | 🌐 Documentation

🎯 Claude Code ⇄ Obsidian Seamless Integration

🔄 Automatic Synchronization System

• Real-time Sync: Instantly reflect changes in .claude/ directory to Obsidian vault
• Bidirectional Integration: Complete integration between Claude Code development and Obsidian knowledge management
• Structured Organization: Systematize knowledge using Obsidian's powerful features

🤖 Automated Metadata Enhancement with YAKE Integration

• Advanced Keyword Extraction: YAKE (Yet Another Keyword Extractor) for unsupervised keyword discovery
• Multi-Language Support: English, Japanese, Spanish, French, German, Italian, Portuguese
• Smart Tagging: AI-powered tag suggestions with confidence scoring
• Evidence-Based Classification: Reliable organization with clear rationale for automated decisions

# Enhanced Metadata Example (Secondary Effect)
type: [prompt, code, concept, resource]           # Content nature
tech: [python, react, fastapi, kubernetes, ...]   # Technology stack
domain: [web-dev, ml, devops, mobile, ...]        # Application domain
team: [backend, frontend, ml-research, devops]    # Team ownership
status: [draft, tested, production, deprecated]   # Lifecycle state
complexity: [beginner, intermediate, advanced]    # Skill level
confidence: [low, medium, high]                   # Content reliability

🏛️ Obsidian-Optimized Vault Structure

obsidian-vault/
├── _system/          # Templates and configuration
├── _attachments/     # Media files
├── inbox/            # Unprocessed content
├── active/           # Work-in-progress content
├── archive/          # Completed/deprecated content
└── knowledge/        # Mature knowledge (main area)

Prerequisites

• uv: Modern Python package manager (includes Python 3.11+ automatically)
  • Installation: Follow the official uv installation guide
  • Quick install: curl -LsSf https://astral.sh/uv/install.sh | sh (Unix/macOS) or powershell -ExecutionPolicy ByPass -c "irm https://astral.sh/uv/install.ps1 | iex" (Windows)
• Python: Not required separately - uv manages Python 3.11+ automatically

🎯 3-Minute Claude Code ⇄ Obsidian Integration Experience

🚀 v0.10.0 YAKE Integration: Advanced keyword extraction with 147 passing tests and 28.25% test coverage ensures production stability.

Experience seamless integration:

# Install CKC
uv pip install claude-knowledge-catalyst

# Initialize in Claude Code project
cd your-claude-project
uv run ckc init

# Connect to Obsidian vault
uv run ckc add my-vault /path/to/obsidian/vault

# Sync .claude/ files with Obsidian
uv run ckc sync

What happens:

• ✅ Seamless Integration: Complete integration between Claude Code development and Obsidian knowledge management
• ✅ Automatic Structuring: Organize .claude/ content with Obsidian-optimized structure
• ✅ Enhanced Metadata: Automatic tagging that reduces manual classification
• ✅ Real-time Sync: Instantly reflect knowledge accumulation during development process

Core Features

🔄 Claude Code ⇄ Obsidian Complete Integration

• Seamless Sync: Automatic bidirectional sync between .claude/ directory and Obsidian vault
• Structured Migration: Optimization and structural enhancement of existing Obsidian vaults
• Dynamic Query Generation: Automatic generation of Obsidian dataview queries
• Knowledge Discovery: Cross-project search of Claude Code development insights within Obsidian

🚀 YAKE Keyword Extraction (New in v0.10.0)

• Unsupervised Learning: Extract keywords without training data
• Multi-Language: Automatic language detection and processing
• Confidence Scoring: Filter high-quality keyword suggestions
• Technical Content: Optimized for technical documentation and code

🔒 Secure CLAUDE.md Sync

• Privacy-First: Section-level filtering for sensitive information
• Configurable Exclusion: Protect API keys, credentials, personal data
• Safe by Default: CLAUDE.md sync disabled unless explicitly enabled

📊 Obsidian Integrated Analytics

• Knowledge Usage Tracking: Analyze knowledge utilization patterns in Claude Code development
• Prompt Effectiveness Measurement: Success rates and improvement suggestions within Obsidian
• Cross-Project Insights: Discover relationships between development insights
• Team Knowledge Sharing: Collaborative knowledge management through Obsidian

🎨 Obsidian-Optimized Templates

• Claude Code Specialized: Obsidian templates for prompts, code, concepts, and resources
• Smart Suggestions: Automatic template selection based on development context
• Evolving Structure: Obsidian vault optimization according to project growth

Quick Start

Installation

# Install from PyPI using uv (recommended)
uv pip install claude-knowledge-catalyst

# Or using pip
pip install claude-knowledge-catalyst

# Or install from source for development
git clone https://github.com/drillan/claude-knowledge-catalyst.git
cd claude-knowledge-catalyst
uv sync --dev

Claude Code Project Integration

# Navigate to Claude Code project
cd your-claude-project

# Initialize CKC (detects .claude/ directory)
uv run ckc init

# Connect to Obsidian vault
uv run ckc add main-vault /path/to/your/obsidian/vault

# Experience automatic analysis of .claude/ content
echo "# Git Useful Commands

## Branch Status Check
\`\`\`bash
git branch -vv
git status --porcelain
\`\`\`" > .claude/git_tips.md

# Verify automated analysis and Obsidian metadata generation
uv run ckc classify .claude/git_tips.md --show-evidence

Existing Obsidian Vault Enhancement

# Enhance existing Obsidian vault for Claude Code integration
uv run ckc migrate --source /existing/obsidian --target /enhanced/vault

# Preview changes
uv run ckc migrate --source /existing/obsidian --target /enhanced/vault --dry-run

Available CLI Commands

🚀 Automated Classification

# Automatic content analysis
uv run ckc classify file.md --show-evidence

# Batch classification
uv run ckc batch-classify .claude/

# Missing metadata detection
uv run ckc scan-missing-metadata

📁 Core Operations

# Zero-config initialization
uv run ckc init

# Vault connection
uv run ckc add vault-name /path/to/obsidian

# State-based synchronization
uv run ckc sync
uv run ckc sync --project "My Project"

# Real-time monitoring
uv run ckc watch

# System status
uv run ckc status

📊 Advanced Analytics

# File analysis with evidence
uv run ckc analyze .claude/my-prompt.md

# Cross-dimensional search
uv run ckc search --tech python --status production
uv run ckc search --team frontend --complexity advanced

# Project insights
uv run ckc project stats my-project

Configuration

CKC uses a YAML configuration file with pure tag-centered settings:

version: "1.0"
project_name: "My AI Project"
auto_sync: true

# Tag-centered architecture
tag_system:
  enabled: true
  multi_dimensional: true
  auto_classification: true
  confidence_threshold: 0.75

# 7-dimensional tag schema
tags:
  type_tags: ["prompt", "code", "concept", "resource"]
  tech_tags: ["python", "javascript", "react", "fastapi"]
  domain_tags: ["web-dev", "machine-learning", "devops"]
  team_tags: ["backend", "frontend", "ml-research"]
  status_tags: ["draft", "tested", "production", "deprecated"]
  complexity_tags: ["beginner", "intermediate", "advanced"]
  confidence_tags: ["low", "medium", "high"]

# Obsidian integration
sync_targets:
  - name: "main-vault"
    type: "obsidian"
    path: "/Users/me/Documents/ObsidianVault"
    enabled: true
    enhance_metadata: true

# Automated features
automation:
  auto_classification: true
  evidence_tracking: true
  natural_language_search: true

# State-based workflow
workflow:
  inbox_pattern: "status:draft"
  active_pattern: "status:tested"
  knowledge_pattern: "status:production"
  archive_pattern: "status:deprecated"

# Security settings
watch:
  include_claude_md: false  # Enable with caution
  claude_md_sections_exclude:
    - "# secrets"
    - "# private"
    - "# api-keys"

Architecture

CKC implements a revolutionary pure tag-centered architecture:

• Cognitive Load Zero: Eliminates category decision fatigue
• 7-Dimensional Classification: Multi-layer tag system for precise organization
• Automated Intelligence: Pattern-matching content understanding
• State-Based Workflow: Organization by lifecycle, not content type
• Dynamic Discovery: Cross-dimensional knowledge search
• Obsidian Enhancement: Transform basic vaults → intelligent systems

Pure Tag-Centered vs Traditional

❌ Traditional Category-Based Problems

├── prompts/          # "Is this a prompt or template?"
├── code/             # "Code snippet or tool?"
├── concepts/         # "Concept or best practice?"
└── misc/             # Catch-all confusion

Issues:

• Decision fatigue: Which category?
• Rigid boundaries: Content doesn't fit neatly
• Poor discoverability: Single-dimension search
• Maintenance overhead: Moving files between categories

✅ Pure Tag-Centered Solution

├── _system/          # System files and templates
├── inbox/            # Unprocessed items (workflow state)
├── active/           # Currently working (activity state)
├── archive/          # Deprecated/old (lifecycle state)
└── knowledge/        # Mature content (90% of files)
    └── Dynamic discovery through enhanced multi-layer tags

Benefits:

• 🧠 Cognitive Load Reduction: No "which category?" decisions
• 🔍 Multi-Dimensional Discovery: Search across tech, domain, team
• 📈 Scalable Organization: Tags evolve with your knowledge
• ⚡ Flexible Workflow: State-based, not content-based organization
• 🔗 Rich Relationships: Multi-project, multi-domain connections

Documentation

• 📖 Documentation - Complete user guide and developer reference
• 🚀 Quick Start - 5-minute Pure Tag-Centered experience
• 👥 User Guide - Practical usage methods
• 🔧 Developer Guide - Developer reference

Try the Revolution

Demo the cognitive transformation:

# Experience tag-centered migration
./demo/tag_centered_demo.sh

# Try automated classification  
./demo/demo.sh

# Multi-team collaboration
./demo/multi_project_demo.sh

Requirements

• Python Runtime: 3.11+ (managed automatically by uv)
• Package Manager: uv (handles Python installation and dependency management)
• Memory: Minimum 512MB, Recommended 2GB for large vaults
• Storage: 10MB for CKC, varies based on vault size
• OS: Windows 10+, macOS 11+, Linux (Ubuntu 20.04+)

Support & Community

• Issues: GitHub Issues
• Discussions: GitHub Discussions
• Documentation: Read the Docs

License

This project is licensed under the MIT License - see the LICENSE file for details.

Contributing

We welcome contributions! Please see our Contributing Guide for details.

---

Welcome to the cognitive revolution!
No more "which category?" decisions - experience pure, discoverable knowledge management.

Built with ❤️ by the Claude community