GitHub - sujankapadia/claude-code-analytics: Analysis tool for Claude Code - automatically capture, search, and analyze your AI development conversations

17 min read Original article ↗

Claude Code Analytics

Tests codecov Python 3.9+ License: MIT

A analysis tool for Claude Code that automatically captures, archives, and analyzes your AI development conversations. Features an interactive dashboard, full-text search, and AI-powered insights across all your sessions.

What It Does

Claude Code Analytics transforms your AI development workflow into actionable insights:

  • Automatically captures every conversation when you exit Claude Code
  • Stores and indexes conversations in a searchable SQLite database
  • Provides an interactive dashboard to explore patterns, tool usage, and decisions
  • Enables AI-powered analysis of your development sessions using 300+ LLM models

Prerequisites

Before installing, you need:

  • Claude Code - The AI coding assistant (this tool captures its conversations)
  • Python 3.9+ - Check with python3 --version
  • jq - JSON processor for installation script
    • macOS: brew install jq
    • Linux: apt-get install jq or yum install jq

Compatibility

Testing Status: This is an alpha release (v0.1.0) that has been tested on:

  • macOS 14.7 (Sonoma)
  • Python 3.9-3.12 (CI tested on Ubuntu and macOS)
  • Claude Code 2.1.x

Platform Support:

  • ✅ macOS: Fully tested and supported
  • ⚠️ Linux: Likely compatible (CI tests pass on Ubuntu) but not extensively tested on user systems
  • ❓ Windows: Untested - hook scripts may require WSL or adaptation

If you encounter platform-specific issues, please report them on GitHub.

Quick Start

1. Install

git clone https://github.com/sujankapadia/claude-code-analytics.git
cd claude-code-analytics
./install.sh

The installer automatically:

  • Installs the Python package and all dependencies
  • Sets up hooks to capture conversations
  • Creates the CLI commands
  • Configures Claude Code settings

2. Import Existing Conversations

This creates the database, imports all existing conversations, and builds the search index.

3. Launch Dashboard

The dashboard opens at http://localhost:8501. Start exploring your conversations!

4. (Optional) Configure AI Analysis

To enable AI-powered analysis features, add an API key to your config file:

# Edit the config file
nano ~/.config/claude-code-analytics/.env

# Add one of these:
OPENROUTER_API_KEY=sk-or-your-key-here  # For 300+ models
GOOGLE_API_KEY=your-key-here             # For Gemini models

Get API keys from OpenRouter or Google AI Studio.

That's it! New conversations will be automatically captured when you exit Claude Code sessions.

⚠️ Important: Transcript Retention

Claude Code automatically removes transcripts for sessions that have been inactive for more than 30 days by default. This helps manage disk space but means older conversations may be deleted before you archive them.

To change this retention period, edit ~/.claude/settings.json and add:

{
  "cleanupPeriodDays": 90
}

Set this to a higher value (e.g., 90, 180, or 365 days) to keep transcripts longer, or set it to 0 to disable automatic cleanup entirely.

Key Features

📊 Interactive Dashboard

The Streamlit-based dashboard is your primary interface for exploring conversations:

  • Session Browser - View, filter, and navigate all your Claude Code sessions with pagination support, session-level activity metrics (active time, text volume), and project-level aggregate totals
  • Conversation Viewer - Terminal-style interface that faithfully recreates your sessions:
    • Inline tool calls and results
    • Role-based filtering (user/assistant)
    • Content search within sessions
    • Token usage display
    • Deep linking to specific messages from search results
  • Analytics Dashboard - Visual insights into your development patterns:
    • Messages and token usage over time
    • Tool usage distribution and error rates
    • Project statistics (sessions, messages, tool uses, activity timeline)
    • Daily activity trends
    • Activity & volume metrics (active time, text volume ratios, per-project breakdown)
  • Full-Text Search - FTS5-powered search across all messages, tool inputs, and tool results:
    • Scope filtering (messages, tool inputs/results)
    • Project and date range filters
    • Highlighted search results with context
    • Direct navigation to matching messages
    • MCP tool usage analysis
  • AI-Powered Analysis - Run sophisticated analysis on any session:
    • Technical decisions extraction
    • Error pattern analysis
    • AI agent usage patterns
    • Custom analysis with your own prompts
    • 300+ model selection via OpenRouter or Gemini

🔍 Search & Discovery

  • Full-text search - Lightning-fast FTS5 search across millions of tokens
  • Deep linking - Search results link directly to specific messages in conversations
  • Advanced filtering - Filter by project, date range, role, tool name
  • MCP tool tracking - Dedicated analytics for MCP server usage
  • Message-level precision - Every tool use is linked to its exact message

💾 Automatic Archiving

  • Hook-based capture - Conversations automatically export on session end
  • Dual-format storage - Raw JSONL for programmatic access, formatted text for reading
  • Project organization - Conversations organized by the project directory they occurred in
  • Incremental imports - Database updates efficiently with only new content
  • Session resumption - Correctly handles resumed sessions and updates

🤖 AI-Powered Analysis

  • 300+ models - Access entire OpenRouter catalog or use Google Gemini directly
  • Curated selection - Quick-select from 13 newest premium models (2025):
    • Budget: Qwen3, Llama 4 Scout, Mistral Small ($0.06-$0.10/1M tokens)
    • Balanced: DeepSeek V3.2, Gemini 3 Flash, Claude Haiku 4.5 ($0.26-$1.75/1M tokens)
    • Premium: Gemini 3 Pro, Claude Sonnet 4.5, Grok 4, Claude Opus 4.5 ($2-$5/1M tokens)
  • Pre-built analysis types:
    • Technical Decisions - Extract decisions, alternatives, and reasoning
    • Error Patterns - Identify recurring issues, root causes, resolutions
    • AI Agent Usage - Understand how you use AI for prototyping and discovery
    • Custom - Write your own analysis prompts
  • Templated prompts - Jinja2-based templates for easy customization
  • Export results - Save analysis as markdown files
  • GitHub Gist publishing - Share analysis with automatic security scanning:
    • Multi-layer scanning (Gitleaks + Regex patterns) for secrets, PII, and sensitive data
    • Blocks publication on CRITICAL/HIGH severity findings
    • Optional session transcript inclusion
    • Public or secret gist visibility
    • Auto-generated README with metadata and traceability

📈 Comprehensive Analytics

  • Token tracking - Input, output, and cache metrics (creation, read, 5m, 1h)
  • Tool usage stats - Track which tools you use most, error rates, session distribution
  • Daily trends - Message volume, token usage, and activity over time
  • Project insights - Compare activity levels across different projects
  • Activity metrics - Active time per session (idle gaps capped at 5 min), text volume ratios (user vs assistant including tool text), per-project breakdowns

Quick Start

1. Install

git clone https://github.com/yourusername/claude-code-analytics.git
cd claude-code-analytics
./install.sh

The installer sets up hooks, creates directories, and configures Claude Code to automatically export conversations.

2. Create Database

# Create database schema
python3 scripts/create_database.py

# Import existing conversations
python3 scripts/import_conversations.py

# Create search index
python3 scripts/create_fts_index.py

3. Launch Dashboard

The dashboard opens at http://localhost:8501. Start exploring your conversations!

4. (Optional) Configure AI Analysis

To use AI-powered analysis features:

# Option 1: OpenRouter (300+ models)
export OPENROUTER_API_KEY="sk-or-your-key-here"

# Option 2: Google Gemini (direct)
export GOOGLE_API_KEY="your-api-key-here"

Get API keys from OpenRouter or Google AI Studio.

Configuration

Claude Code Analytics uses a centralized configuration system with smart defaults. All settings are stored in ~/.config/claude-code-analytics/.env.

Configuration File Location

The configuration file is automatically created during installation at:

~/.config/claude-code-analytics/.env

This location follows the XDG Base Directory specification.

Configuration Variables

The .env file supports all standard environment variable syntax, including variable interpolation using ${VAR} syntax.

Data Directories

# Base directory for all conversation data
# All other paths default to subdirectories of this unless explicitly overridden
CLAUDE_CONVERSATIONS_DIR=~/claude-conversations

# Optional: Override specific directories
ANALYSIS_OUTPUT_DIR=${CLAUDE_CONVERSATIONS_DIR}/analyses
DATABASE_PATH=${CLAUDE_CONVERSATIONS_DIR}/conversations.db

Pagination Settings

# Number of messages before enabling pagination in conversation viewer
PAGINATION_THRESHOLD=500

# Number of messages to show per page when paginated
MESSAGES_PER_PAGE=100

Search Configuration

# Number of search results to show per page
SEARCH_RESULTS_PER_PAGE=10

Display Settings

# Maximum length of tool results to display (characters)
# Results longer than this will be truncated
TOOL_RESULT_MAX_LENGTH=2000

Debug Logging

# Location of export hook debug log
CLAUDE_EXPORT_DEBUG_LOG=~/.claude/export-debug.log

LLM API Configuration (Required for AI Analysis)

# OpenRouter API key (access 300+ models)
# Get your key from: https://openrouter.ai/keys
OPENROUTER_API_KEY=sk-or-your-key-here

# Default model for CLI analysis (when --model not specified)
OPENROUTER_MODEL=deepseek/deepseek-v3.2

# Google Gemini API key (alternative to OpenRouter)
# Get your key from: https://aistudio.google.com/app/apikey
GOOGLE_API_KEY=your-api-key-here

GitHub Integration (Optional - for Gist Publishing)

# GitHub Personal Access Token for publishing analysis as gists
# Get your token from: https://github.com/settings/tokens/new
# Required scope: 'gist' (create gists)
GITHUB_TOKEN=ghp_your_token_here

Required vs Optional Settings

Required for basic features (browse, search, analytics):

  • None! The defaults work out of the box.

Required for AI analysis features:

  • Either OPENROUTER_API_KEY or GOOGLE_API_KEY

Required for GitHub Gist publishing:

  • GITHUB_TOKEN - Personal Access Token with gist scope

Optional customization:

  • All other settings have sensible defaults
  • Customize data directories if you want conversations stored elsewhere
  • Adjust pagination/display settings for personal preferences

Customizing Configuration

Edit the configuration file to customize settings:

# Open configuration file in your default editor
${EDITOR:-nano} ~/.config/claude-code-analytics/.env

Example customizations:

# Store conversations on external drive
CLAUDE_CONVERSATIONS_DIR=~/Dropbox/claude-conversations

# Increase pagination threshold for faster sessions
PAGINATION_THRESHOLD=1000
MESSAGES_PER_PAGE=200

# Show more search results per page
SEARCH_RESULTS_PER_PAGE=25

# Show longer tool results without truncation
TOOL_RESULT_MAX_LENGTH=5000

Variable Interpolation

The configuration system supports referencing other variables using ${VAR} syntax:

# Set base directory
CLAUDE_CONVERSATIONS_DIR=~/my-custom-location

# Derived paths automatically use the base
ANALYSIS_OUTPUT_DIR=${CLAUDE_CONVERSATIONS_DIR}/analyses
DATABASE_PATH=${CLAUDE_CONVERSATIONS_DIR}/conversations.db

Environment Variables

You can override configuration file settings using environment variables:

# Temporary override for this session
export OPENROUTER_MODEL="anthropic/claude-sonnet-4.5"
./run_dashboard.sh

# Permanent override in shell profile (~/.bashrc, ~/.zshrc)
echo 'export OPENROUTER_API_KEY="sk-or-your-key"' >> ~/.zshrc

Configuration Reference

For a complete list of all configuration variables with documentation, see .env.example in the repository.

Using the Dashboard

Browse Sessions

The Browse Sessions page shows all your conversations:

  • Filter by project, date range, or minimum message count
  • Sort by date or activity level
  • Pagination for large conversation histories
  • Click any session to view full conversation
  • Session activity metrics: active time, message counts, text volume ratios
  • Project-level totals: aggregate active time, average per session, total text volume

Search Conversations

The Search page provides powerful full-text search:

  • Search across messages, tool inputs, or tool results
  • Filter by project, date range, or specific tools
  • View MCP tool usage statistics
  • Click search results to jump directly to matching messages in context

View Analytics

The Analytics Dashboard provides visual insights:

  • Tool Usage - Distribution of top 10 tools, error rates, and session usage
  • Daily Activity - Messages, tokens, and sessions over time (configurable time range)
  • Token Usage - Input vs output tokens with stacked area chart
  • Project Statistics - Sessions, messages, tool uses, and activity timeline per project (sorted by message volume)
  • Activity & Volume Metrics - Total active time, average per session, text volume (user vs assistant with ratios), per-project breakdown table and bar chart

Run AI Analysis

The AI Analysis page lets you analyze sessions with LLMs:

  1. Select a session from the dropdown
  2. Choose analysis type or write custom prompt
  3. Select model (browse 300+ options or pick from curated list)
  4. Adjust temperature (default: 0.1 for deterministic analysis)
  5. Run analysis and optionally export to markdown

Publish to GitHub Gists

After running an analysis, you can publish your results as a GitHub Gist with automatic security scanning:

  1. Configure GitHub Token - Add GITHUB_TOKEN to ~/.config/claude-code-analytics/.env
  2. Run Analysis - Complete an analysis on any session
  3. Configure Gist Options:
    • Choose visibility (Secret/unlisted or Public)
    • Optionally include raw session transcript
    • Customize gist description
  4. Scan & Publish - Click "Scan & Publish to Gist" button
  5. Review Results:
    • ✅ Safe: Gist is published with URL and scan summary
    • ❌ Blocked: Security findings prevent publication (review and remove sensitive data)

Security Scanning:

  • Gitleaks - Detects 350+ secret patterns (API keys, tokens, credentials)
  • Regex Patterns - Catches PII (emails, phone numbers, SSNs, credit cards)
  • Severity Levels:
    • CRITICAL/HIGH → Blocks publication
    • MEDIUM/LOW → Informational warnings only

The gist includes:

  • Analysis result with full metadata and traceability
  • Optional session transcript
  • Auto-generated README with tool attribution

Advanced Usage

CLI Tools

The installation provides several CLI commands for working with your conversations:

Search Conversations

claude-code-search "error handling"

Run Analysis from CLI

# Analyze technical decisions
claude-code-analyze <session-id> --type=decisions

# Specify model and save output
claude-code-analyze <session-id> \
  --type=errors \
  --model=anthropic/claude-sonnet-4.5 \
  --output=analysis.md

# Custom analysis
claude-code-analyze <session-id> \
  --type=custom \
  --prompt="Summarize key technical insights"

Popular models:

  • deepseek/deepseek-v3.2 - Best balance (default, $0.26/1M)
  • anthropic/claude-sonnet-4.5 - Highest quality ($3.00/1M)
  • openai/gpt-5.2-chat - Latest GPT ($1.75/1M)
  • google/gemini-3-flash-preview - 1M context window ($0.50/1M)

Import New Conversations

Run the import command anytime to update the database with new conversations:

The command automatically:

  • Detects existing sessions
  • Imports only new messages
  • Updates session metadata (end times, message counts)
  • Preserves all existing data with zero duplicates
  • Works efficiently on active or completed sessions

Using Python Scripts Directly

If you prefer to use the Python scripts directly instead of CLI commands:

# Search
python3 scripts/search_fts.py "error handling"

# Analyze
python3 scripts/analyze_session.py <session-id> --type=decisions

# Import
python3 scripts/import_conversations.py

Manual Export

Convert JSONL transcripts to readable text format:

~/.claude/scripts/pretty-print-transcript.py /path/to/transcript.jsonl output.txt

# Or via stdin/stdout
cat transcript.jsonl | ~/.claude/scripts/pretty-print-transcript.py > output.txt

Custom Analysis Prompts

Create custom analysis templates in prompts/:

  1. Create a new .md file with your Jinja2 template
  2. Add metadata to prompts/metadata.yaml
  3. Use from dashboard or CLI with --type=your_template_name

See prompts/README.md for detailed instructions.

How It Works

Architecture

Claude Code Session
       ↓
SessionEnd Hook (export-conversation.sh)
       ↓
~/claude-conversations/
  ├── project-name/
  │   ├── session-YYYYMMDD-HHMMSS.jsonl  (raw data)
  │   └── session-YYYYMMDD-HHMMSS.txt    (readable format)
       ↓
Import Script (import_conversations.py)
       ↓
SQLite Database (conversations.db)
  ├── projects
  ├── sessions
  ├── messages
  ├── tool_uses
  └── fts_messages (full-text search index)
       ↓
Streamlit Dashboard
  ├── Browse & filter sessions
  ├── Search conversations
  ├── View analytics
  └── Run AI analysis

Hook System

The export-conversation.sh hook runs automatically when you exit Claude Code:

  1. Receives current working directory and transcript path
  2. Finds the most recent transcript (handles session resumption)
  3. Creates project-specific directory structure
  4. Copies JSONL file with timestamp
  5. Generates human-readable text version
  6. Logs to ~/.claude/export-debug.log

Database Schema

The SQLite database uses a normalized schema:

  • projects - Unique project directories
  • sessions - Conversation sessions with metadata
  • messages - Individual messages with token tracking
  • tool_uses - Tool calls linked to messages via message_index
  • fts_messages - FTS5 full-text search index
  • Views - Pre-aggregated statistics for performance

See the technical documentation in docs/technical/ for more details.

Project Identification

Important: Claude Code uses the working directory path as the project identifier, not a separate project name or git repository name.

When you run claude-code in a directory like /Users/username/dev/my-project, Claude Code:

  1. Encodes the path by replacing / with -:

    • /Users/username/dev/my-project → -Users-username-dev-my-project

  2. Stores sessions in ~/.claude/projects/{encoded-path}/:

    ~/.claude/projects/
  └── -Users-username-dev-my-project/
      ├── session-uuid-1.jsonl
      ├── session-uuid-2.jsonl
      └── ...

  3. Analytics tool decodes the path back to display in the dashboard:

    • project_id (database): -Users-username-dev-my-project (as stored by Claude Code)
    • project_name (display): /Users/username/dev/my-project (decoded for readability)

This means:

  • Same directory = same project across all your sessions
  • Different paths to the same repo = different projects (e.g., /home/user/project vs /home/user/repos/project)
  • Subdirectories are treated as separate projects (e.g., /project vs /project/subdir)

The analytics dashboard groups all conversations by these directory paths, so you'll see all your work for a specific codebase location together.

File Organization

Exported conversations are organized by project:

~/claude-conversations/
├── project-name-1/
│   ├── session-20250113-143022.jsonl
│   ├── session-20250113-143022.txt
│   ├── session-20250113-151430.jsonl
│   └── session-20250113-151430.txt
├── project-name-2/
│   └── session-20250114-091500.jsonl
└── conversations.db

Readable Text Format

The generated .txt files provide a clean, readable format:

═══════════════════════════════════════════════════════════════
USER (2025-01-13 14:30:22)
───────────────────────────────────────────────────────────────
Can you help me fix the bug in the authentication module?

═══════════════════════════════════════════════════════════════
CLAUDE (2025-01-13 14:30:24)
───────────────────────────────────────────────────────────────
I'll help you fix the authentication bug. Let me first examine the
authentication module to understand the issue.

[Tool: Read]
$ Read file_path=/path/to/auth.js

[Tool Result]
1  function authenticate(user, password) {
2    if (user && password) {
3      return true;
4    }
5  }

Manual Installation

If you need to install manually or want to understand what the installer does:

1. Create directories

mkdir -p ~/.claude/scripts
mkdir -p ~/claude-conversations

2. Copy scripts

cp hooks/export-conversation.sh ~/.claude/scripts/
cp scripts/pretty-print-transcript.py ~/.claude/scripts/
chmod +x ~/.claude/scripts/export-conversation.sh
chmod +x ~/.claude/scripts/pretty-print-transcript.py

3. Configure hook

Add to ~/.claude/settings.json:

{
  "hooks": {
    "SessionEnd": [
      {
        "matcher": "",
        "hooks": [
          {
            "type": "command",
            "command": "bash ~/.claude/scripts/export-conversation.sh"
          }
        ]
      }
    ]
  }
}

If you have existing hooks, merge the SessionEnd entry into your existing hooks object.

4. Install Python package and dependencies

# From the repository directory
pip install -e .

# This installs the package and all dependencies, and creates the CLI commands

Troubleshooting

Conversations not exporting

Check the debug log:

cat ~/.claude/export-debug.log

Common issues:

  • Hook not configured in ~/.claude/settings.json
  • Scripts not executable (chmod +x)
  • Incorrect paths in settings

Permission errors

Ensure directories are writable:

chmod 755 ~/claude-conversations
chmod 755 ~/.claude/scripts

Import errors

If database import fails:

  • Verify JSONL files exist in ~/claude-conversations/
  • Check file permissions
  • Ensure Python 3.9+ is installed
  • Run with verbose output: claude-code-import -v

Dashboard not launching

  • Install dependencies: pip install streamlit pandas altair
  • Check port 8501 is available
  • Try alternate port: streamlit run streamlit_app/app.py --server.port=8502

Project Structure

claude-code-analytics/
├── hooks/
│   └── export-conversation.sh       # SessionEnd hook for auto-export
├── scripts/
│   ├── pretty-print-transcript.py   # Convert JSONL to readable text
│   ├── create_database.py           # Create SQLite database schema
│   ├── import_conversations.py      # Import conversations to database
│   ├── create_fts_index.py          # Create full-text search index
│   ├── search_fts.py                # CLI search tool
│   └── analyze_session.py           # CLI analysis tool
├── streamlit_app/
│   ├── app.py                       # Dashboard entry point
│   ├── models/                      # Pydantic data models
│   │   └── __init__.py
│   ├── services/                    # Business logic layer
│   │   ├── database_service.py
│   │   └── format_utils.py
│   └── pages/                       # Dashboard pages
│       ├── browser.py               # Session browser
│       ├── conversation.py          # Conversation viewer
│       ├── search.py                # Full-text search
│       ├── analytics.py             # Analytics dashboard
│       └── ai_analysis.py           # AI-powered analysis
├── prompts/                         # Analysis prompt templates
│   ├── metadata.yaml                # Analysis type definitions
│   ├── decisions.md                 # Technical decisions prompt
│   ├── errors.md                    # Error patterns prompt
│   └── agent_usage.md               # AI agent usage analysis
├── docs/                            # Documentation
│   ├── index.html                   # Landing page (GitHub Pages)
│   └── technical/                   # Technical documentation
│       ├── agent-knowledge-retention.md
│       ├── deep-linking-implementation.md
│       ├── mcp-server-analytics.md
│       └── search-feature-requirements.md
├── install.sh                       # Automated installer
├── run_dashboard.sh                 # Dashboard launcher
└── README.md                        # This file

Documentation

Development

Setting Up Development Environment

If you're contributing to Claude Code Analytics or modifying the code, we use pre-commit hooks to ensure code quality:

1. Install pre-commit

2. Install the git hooks

3. (Optional) Run against all files

pre-commit run --all-files

The hooks will now run automatically before each commit.

Code Quality Tools

The pre-commit configuration includes:

  • Black - Automatic code formatting (100 character line length)
  • Ruff - Fast Python linter with auto-fix capabilities
    • Checks: pycodestyle, pyflakes, isort, flake8-bugbear, pyupgrade, and more
    • Automatically modernizes type hints (e.g., typing.Dict → dict)
    • Improves exception handling patterns
  • Bandit - Security linting to detect potential vulnerabilities
  • Mypy - Static type checking for type safety
  • Standard checks - Trailing whitespace, end-of-file, YAML/JSON validation, large files, merge conflicts, private keys

Manual Code Quality Checks

If you need to run checks manually without committing:

# Run all hooks
pre-commit run --all-files

# Run specific hook
pre-commit run black --all-files
pre-commit run ruff --all-files

# Skip hooks for a specific commit (not recommended)
git commit --no-verify

Configuration

All tool configurations are in pyproject.toml:

  • Black: 100 character line length
  • Ruff: Comprehensive rule set with modern Python practices
  • Bandit: Excludes test files, configured for security-critical patterns
  • Mypy: Python 3.9+ target with reasonable strictness

Logging Standards

Production scripts use Python's logging module for consistent output. See claude_code_analytics/scripts/LOGGING.md for conventions:

import logging

logging.basicConfig(
    level=logging.INFO,
    format='%(levelname)s: %(message)s'
)
logger = logging.getLogger(__name__)

# Use logger methods instead of print()
logger.info("Processing completed")
logger.error("Database not found")
logger.warning("Interrupted by user")

Future Roadmap

  • Vector embeddings - Semantic search across conversations
  • Cost tracking - Monitor LLM API costs per analysis
  • Model comparison - A/B test analysis quality across models
  • Export formats - HTML, PDF conversation exports
  • Real-time analysis - Analyze conversations as they happen
  • Cloud sync - Optional backup to cloud storage
  • Presidio integration - Advanced PII detection with entity recognition

Contributing

Contributions are welcome! Feel free to:

  • Report bugs or request features via GitHub Issues
  • Submit pull requests
  • Share your custom analysis prompts
  • Suggest new analytics visualizations

License

MIT License - Use and modify freely.

Resources