@okrapdf/cli

Command-line interface for OkraPDF - extract tables from PDFs.

Installation

npm install -g @okrapdf/cli

Or use with npx:

npx @okrapdf/cli jobs list

Quick Start

# Create a .env file with your API key
echo "OKRA_API_KEY=okra_xxxxxxxxxxxx" > .env

# Or create ~/.okra for global config
echo "OKRA_API_KEY=okra_xxxxxxxxxxxx" > ~/.okra

# List your documents
okra docs list

# List extraction jobs  
okra jobs list

# Extract tables from a PDF
okra extract invoice.pdf -o json

Get your API key from okrapdf.com/settings/api-keys.

Configuration

The CLI looks for your API key in this order:

1. OKRA_API_KEY environment variable
2. .env file in current directory
3. .okra file in current directory
4. ~/.okra file in home directory
5. ~/.config/okrapdf/config.json

Example .env or .okra file:

OKRA_API_KEY=okra_xxxxxxxxxxxx
OKRA_BASE_URL=https://okrapdf.com

Commands

Documents

okra docs list               # List all documents
okra docs upload <file>      # Upload a PDF
okra docs get <uuid>         # Get document details
okra docs delete <uuid>      # Delete a document

Jobs

okra jobs list               # List extraction jobs
okra jobs create <file>      # Create extraction job
okra jobs get <job-id>       # Get job status
okra jobs wait <job-id>      # Wait for completion
okra jobs results <job-id>   # Get extraction results

Tables

okra tables list <doc-uuid>  # List extracted tables
okra tables get <table-id>   # Get table content
okra tables export <id>      # Export to CSV/JSON

Shortcuts

okra extract <file>          # Upload + extract + wait (all-in-one)

Interactive

okra chat <job-id>           # Interactive document chat
okra watch [job-ids...]      # Monitor live sessions in browser

Introspection

okra schema [command ...]    # Machine-readable command schema
okra exit-codes              # Stable exit codes for automation

Output Formats

All commands support -o, --output:

• table (default) - Human-readable tables
• json - Machine-readable JSON
• csv - CSV format

# JSON for scripting
okra jobs list -o json | jq '.[].id'

# Quiet mode for piping
okra extract doc.pdf -o json -q > results.json

Environment Variables

| Variable | Description | |----------|-------------| | OKRA_API_KEY | API key (required) | | OKRA_BASE_URL | Custom API URL (for self-hosted) | | OKRA_OUTPUT_FORMAT | Default output format | | OKRA_VERBOSE | Enable verbose diagnostics when set to 1 |

Examples

Batch Processing

for pdf in *.pdf; do
  okra extract "$pdf" -o json > "${pdf%.pdf}.json"
done

CI/CD Integration

# Extract and check for tables
RESULT=$(okra extract report.pdf -o json -q)
TABLE_COUNT=$(echo "$RESULT" | jq '.tables | length')
echo "Found $TABLE_COUNT tables"

Exit Codes

| Code | Meaning | |------|---------| | 0 | Success | | 1 | General error | | 2 | Invalid arguments | | 3 | Authentication error | | 4 | Resource not found | | 5 | Rate limited |

Development

git clone https://github.com/steventsao/okrapdf
cd okrapdf/packages/cli
npm install
npm run build
npm link
okra --help

License

MIT

Links

• OkraPDF
• API Docs
• GitHub