crdbextract

Tools for extracting data from Chrome's IndexedDB storage and Claude Code sessions.

Features

• Chrome IndexedDB Extraction: Extract data from Chrome's LevelDB-based IndexedDB storage
• Multi-Profile Support: Work with multiple Chrome profiles
• Safe Extraction: Copy databases before reading to avoid Chrome file locks
• Claude Code Sessions: Extract conversation history from Claude Code's local storage

Installation

# Clone the repository
git clone https://github.com/mtaran/crdbextract.git
cd crdbextract

# Create virtual environment
python3 -m venv venv
source venv/bin/activate

# Install dependencies
pip install -r requirements.txt

Usage

Chrome IndexedDB Extraction

# List all IndexedDB databases
python extract_indexeddb.py --list

# List Chrome profiles
python extract_indexeddb.py --list-profiles

# Extract a specific database
python extract_indexeddb.py --path "/path/to/https_example.com_0.indexeddb.leveldb" --pretty

# Extract with safe copy (avoids Chrome locks)
python extract_indexeddb.py --path "/path/to/db" --safe-copy --output data.json

# Extract from a specific profile
python extract_indexeddb.py --profile "Profile 1" --list

# Extract all databases
python extract_indexeddb.py --all --output all_data.json

Claude Code Session Extraction

Claude Code stores conversation sessions locally at ~/.claude/projects/ as JSONL files.

# List all Claude Code session files
python extract_claude_sessions.py --list

# Copy all session files to a directory
python extract_claude_sessions.py --output ./sessions/

# Copy sessions from a specific project
python extract_claude_sessions.py --project /path/to/project --output ./sessions/

Files are copied with a project prefix to avoid name collisions (e.g., -Users-mtaran-myproject_session-id.jsonl).

Chrome IndexedDB Location

Platform	Path
macOS	~/Library/Application Support/Google/Chrome/Default/IndexedDB/
Windows	%LOCALAPPDATA%\Google\Chrome\User Data\Default\IndexedDB\
Linux	~/.config/google-chrome/Default/IndexedDB/

Each origin gets a <origin>_0.indexeddb.leveldb directory containing LevelDB files.

How It Works

Chrome stores IndexedDB data in LevelDB format with:

• V8/Blink serialization for JavaScript values
• Snappy compression for data blocks
• Write-ahead logging for durability

This tool uses ccl_chromium_reader to parse the LevelDB format and deserialize the stored values.

Output Format

IndexedDB Output

{
  "path": "/path/to/database.indexeddb.leveldb",
  "origin": "https_example.com_0",
  "databases": [
    {
      "name": "myDatabase",
      "id": 1,
      "object_stores": [
        {
          "name": "myStore",
          "records": [
            {"key": "key1", "value": {"data": "value"}}
          ]
        }
      ]
    }
  ]
}

Notes

• Chrome must be closed for reliable extraction, or use --safe-copy
• Recent writes may be in WAL (.log files) not yet compacted to .ldb files
• Some complex JavaScript objects (Blobs, Files) may not fully deserialize

Credits

• ccl_chromium_reader - LevelDB and IndexedDB parsing
• dfindexeddb - Alternative Google tool for forensic analysis

License

MIT