CLI Reference¶
Complete command-line interface reference for DrGPT with all options, flags, and usage patterns.
Synopsis¶
drgpt [OPTIONS] [PROMPT]
Basic Usage¶
# Simple query
drgpt "Your question here"
# Use specific mode
drgpt --code "Generate code"
drgpt --shell "System command"
drgpt --interface
drgpt --editor
Positional Arguments¶
PROMPT¶
The question or request to send to the AI.
Type: String
Required: No (unless using standard mode)
Default: None
# Basic prompt
drgpt "Explain machine learning"
# Multi-word prompts (quotes recommended)
drgpt "How do I optimize database queries for large datasets?"
Mode Options¶
--code¶
Generate pure code without explanations.
Type: Flag (no arguments)
Default: False
# Generate Python function
drgpt --code "Create a function to sort a list"
# Generate with specific language
drgpt --code "Create a React component for user authentication"
--shell¶
Generate system administration commands with interactive execution.
Type: Flag (no arguments)
Default: False
# Generate system command
drgpt --shell "Find all Python files larger than 1MB"
# Package management
drgpt --shell "Install Docker on Ubuntu"
--interface¶
Start interactive AI chat mode.
Type: Flag (no arguments)
Default: False
# Start interactive session
drgpt --interface
--editor¶
Open vi text editor for prompt composition.
Type: Flag (no arguments)
Default: False
# Open editor for complex prompts
drgpt --editor
Provider Options¶
--provider PROVIDER¶
Specify AI provider to use.
Type: String
Choices:
openai,anthropic,googleDefault: From configuration or
openai
# Use OpenAI
drgpt --provider openai "Your question"
# Use Anthropic
drgpt --provider anthropic "Your question"
# Use Google AI
drgpt --provider google "Your question"
--model MODEL¶
Specify AI model to use.
Type: String
Default: Provider-specific default
OpenAI Models:
- gpt-4
- gpt-4o-mini (default)
- gpt-3.5-turbo
Anthropic Models:
- claude-3-opus-20240229
- claude-3-sonnet-20240229 (default)
- claude-3-haiku-20240307
Google Models:
- gemini-pro (default)
# Specific model
drgpt --provider openai --model gpt-4 "Complex reasoning task"
# Cost-effective model
drgpt --provider openai --model gpt-4o-mini "Simple question"
--list-providers¶
List all available providers and models.
Type: Flag (no arguments)
# Show all providers and models
drgpt --list-providers
--api-key API_KEY¶
Set API key for the current provider.
Type: String
Security: Key is stored securely in configuration
# Set OpenAI key
drgpt --provider openai --api-key "sk-your-key-here"
# Set Anthropic key
drgpt --provider anthropic --api-key "sk-ant-your-key-here"
Output Options¶
--output FILE¶
Save response to file.
Type: String (file path)
Default: None (output to terminal)
# Save to file
drgpt --output response.md "Explain Docker architecture"
# Save with specific format
drgpt --code --output code.py "Create a web scraper"
--no-streaming¶
Disable real-time response streaming.
Type: Flag (no arguments)
Default: False (streaming enabled)
# Get complete response at once
drgpt --no-streaming "Explain quantum computing"
--no-markdown¶
Disable markdown formatting in output.
Type: Flag (no arguments)
Default: False (markdown enabled)
# Plain text output
drgpt --no-markdown "Simple explanation"
# Useful for scripts
RESPONSE=$(drgpt --no-markdown "Brief answer")
Configuration Options¶
--status¶
Show current configuration and provider status.
Type: Flag (no arguments)
# Show current settings
drgpt --status
--set-default¶
Set current provider and model as default.
Type: Flag (no arguments)
Requires:
--providerand optionally--model
# Set OpenAI as default
drgpt --provider openai --model gpt-4 --set-default
Advanced Options¶
--temperature TEMP¶
Control response creativity and randomness.
Type: Float
Range: 0.0 to 1.0
Default: 0.7
# More focused responses
drgpt --temperature 0.1 "Factual explanation"
# More creative responses
drgpt --temperature 0.9 "Creative writing task"
--max-tokens TOKENS¶
Maximum number of tokens in response.
Type: Integer
Default: Provider-specific default
# Shorter responses
drgpt --max-tokens 500 "Brief summary"
# Longer responses
drgpt --max-tokens 2000 "Detailed analysis"
--timeout SECONDS¶
Request timeout in seconds.
Type: Integer
Default: 30
# Longer timeout for complex requests
drgpt --timeout 60 "Complex analysis task"
Debug and Information Options¶
--debug¶
Enable debug mode with detailed logging.
Type: Flag (no arguments)
Default: False
# Show debug information
drgpt --debug "Your question"
--verbose¶
Enable verbose output.
Type: Flag (no arguments)
Default: False
# Verbose logging
drgpt --verbose "Your question"
--version¶
Show DrGPT version information.
Type: Flag (no arguments)
# Show version
drgpt --version
--help¶
Show help message and usage information.
Type: Flag (no arguments)
# Show help
drgpt --help
Configuration File Options¶
--config CONFIG_FILE¶
Use specific configuration file.
Type: String (file path)
Default:
~/.config/drgpt/config
# Use custom config
drgpt --config ~/my-drgpt-config "Your question"
--no-config¶
Ignore configuration file and use defaults.
Type: Flag (no arguments)
# Use only command-line options
drgpt --no-config --provider openai "Your question"
Environment Variables¶
DrGPT respects these environment variables:
API Keys¶
OPENAI_API_KEY # OpenAI API key
ANTHROPIC_API_KEY # Anthropic API key
GOOGLE_API_KEY # Google AI API key
Default Settings¶
DRGPT_DEFAULT_PROVIDER # Default provider
DRGPT_DEFAULT_MODEL # Default model
DRGPT_NO_STREAMING # Disable streaming (true/false)
DRGPT_NO_MARKDOWN # Disable markdown (true/false)
Advanced Settings¶
DRGPT_CONFIG_DIR # Configuration directory
DRGPT_TIMEOUT # Default timeout
DRGPT_MAX_TOKENS # Default max tokens
DRGPT_TEMPERATURE # Default temperature
Exit Codes¶
DrGPT uses these exit codes:
0 # Success
1 # General error
2 # Configuration error
3 # Network error
4 # API error
5 # User interruption (Ctrl+C)
Usage Examples¶
Basic Queries¶
# Simple question
drgpt "What is Docker?"
# Complex question with context
drgpt "How do I implement JWT authentication in a Node.js API with proper security practices?"
Provider and Model Selection¶
# Use specific provider
drgpt --provider anthropic "Analyze this philosophical argument"
# Use specific model for complex task
drgpt --provider openai --model gpt-4 "Design a distributed system architecture"
# Cost-effective model for simple tasks
drgpt --provider openai --model gpt-4o-mini "Explain basic concepts"
Mode-Specific Examples¶
# Code generation
drgpt --code "Create a Python class for handling HTTP requests"
# System administration
drgpt --shell "Set up SSL certificate for nginx"
# Complex prompt composition
drgpt --editor
# Interactive session
drgpt --interface
Output Control¶
# Save to file
drgpt --output analysis.md "Analyze current AI trends"
# Plain text for scripting
RESULT=$(drgpt --no-markdown "Brief answer")
# Immediate formatted output
drgpt --no-streaming "Detailed explanation"
Configuration Management¶
# Check current settings
drgpt --status
# Set new defaults
drgpt --provider anthropic --model claude-3-sonnet-20240229 --set-default
# Use custom configuration
drgpt --config ~/.drgpt-work-config "Work-related question"
Troubleshooting Commands¶
# Debug mode
drgpt --debug "Your question"
# Test provider connectivity
drgpt --provider openai --test
# Verbose output
drgpt --verbose "Your question"
Combining Options¶
Multiple options can be combined:
# Code generation with specific provider and output
drgpt --provider anthropic --code --output webapp.py "Create a Flask web application"
# Shell command with debug and custom timeout
drgpt --shell --debug --timeout 60 "Complex system maintenance task"
# Interactive mode with specific provider
drgpt --interface --provider openai --model gpt-4
Option Precedence¶
When the same setting is specified in multiple places, DrGPT uses this precedence order:
Command-line options (highest precedence)
Environment variables
Configuration file
Built-in defaults (lowest precedence)
Example:
# Environment variable
export DRGPT_DEFAULT_PROVIDER="anthropic"
# Command-line override (takes precedence)
drgpt --provider openai "Your question" # Uses OpenAI, not Anthropic
Shell Integration¶
DrGPT works well with shell features:
Pipes¶
# Pipe file content
cat error.log | drgpt "Analyze this error log"
# Chain with other commands
drgpt --no-markdown "Brief answer" | grep "keyword"
Command Substitution¶
# Use DrGPT output in variables
EXPLANATION=$(drgpt --no-markdown "Explain $TERM")
# Use in other commands
echo "AI says: $(drgpt --no-markdown 'Brief answer')"
Scripting¶
#!/bin/bash
# Script example
# Check if DrGPT is available
if ! command -v drgpt &> /dev/null; then
echo "DrGPT not found"
exit 1
fi
# Get AI assistance
ADVICE=$(drgpt --no-markdown --provider openai "How to optimize $1")
echo "Optimization advice: $ADVICE"
See Also¶
Configuration - Configuration file format and options
Standard Mode - Detailed mode documentation
AI Providers - Provider-specific information
Troubleshooting - Common issues and solutions