nlp2cmd

Examples Guide

This guide provides a comprehensive overview of all NLP2CMD examples, organized by use case and complexity.

Documentation Hub - Entry point for all docs
User Guide - Complete usage tutorial
CLI Reference - Command line usage
Python API Guide - Programmatic usage

🗂️ Examples Structure

examples/
├── thermodynamic_example.py     # Standalone thermodynamic demo
├── use_cases/                    # Real-world scenarios
│   ├── shell_commands_demo.sh    # Complete CLI examples
│   ├── simple_demo_examples.py   # Python API + Shell concepts
│   ├── complete_python_shell_examples.py # Full Python API
│   ├── dsl_commands_demo.py      # Direct DSL generation
│   ├── devops_automation.py      # DevOps workflows
│   ├── data_science_ml.py        # Data science workflows
│   ├── drug_discovery.py          # Drug discovery workflows
│   ├── healthcare.py             # Healthcare applications
│   ├── finance_trading.py         # Financial operations
│   ├── smart_cities.py           # Urban management
│   └── ...
├── architecture/                 # System architecture demos
│   └── end_to_end_demo.py       # Complete workflow
├── sql/                         # SQL examples
│   ├── basic_sql.py            # Simple queries
│   ├── advanced_sql.py         # Complex queries
│   └── sql_workflows.py        # SQL pipelines
├── shell/                       # Shell examples
│   ├── basic_shell.py          # Common operations
│   ├── feedback_loop.py        # Interactive feedback
│   └── environment_analysis.py # System analysis
├── docker/                      # Docker examples
│   ├── basic_docker.py         # Container operations
│   └── file_repair.py          # Dockerfile repair
├── kubernetes/                  # Kubernetes examples
│   └── basic_kubernetes.py     # K8s operations
├── pipelines/                   # Pipeline examples
│   ├── log_analysis.py         # Log processing
│   └── infrastructure_health.py # System monitoring
└── validation/                  # Validation examples
    └── config_validation.py    # File validation

🚀 Quick Start Examples

1. CLI Usage Examples

File: examples/use_cases/shell_commands_demo.sh

The fastest way to see NLP2CMD in action:

# Run the complete CLI demo
./examples/use_cases/shell_commands_demo.sh

# Individual examples
nlp2cmd --query "Pokaż użytkowników"
nlp2cmd --dsl docker --query "Pokaż wszystkie kontenery"
nlp2cmd --dsl shell --query "Znajdź pliki .log większe niż 10MB"
nlp2cmd analyze-env

What you’ll learn:

Basic CLI syntax
Different DSL types
Environment analysis
File validation

2. Python API Concepts

File: examples/use_cases/simple_demo_examples.py

Conceptual overview without installation requirements:

python3 examples/use_cases/simple_demo_examples.py

What you’ll learn:

Python API concepts
Shell command patterns
Mixed workflow approaches
Real-world use cases

3. Complete Python API

File: examples/use_cases/complete_python_shell_examples.py

Full Python API with actual functionality:

from nlp2cmd.generation import HybridThermodynamicGenerator

generator = HybridThermodynamicGenerator()

# DSL generation
result = await generator.generate("Pokaż użytkowników")

# Thermodynamic optimization
result = await generator.generate("Zoptymalizuj przydzielanie zasobów")

What you’ll learn:

HybridThermodynamicGenerator usage
DSL vs thermodynamic routing
Context-aware queries
Performance patterns

4. Thermodynamic Computing Demo

File: examples/thermodynamic_example.py

Focused walkthrough of Langevin sampling, energy estimation, and routing:

python3 examples/thermodynamic_example.py

Companion example: examples/use_cases/drug_discovery.py for molecule optimization.

📚 Domain-Specific Examples

SQL Examples

Basic SQL (examples/sql/basic_sql.py)

from nlp2cmd import NLP2CMD, SQLAdapter

nlp = NLP2CMD(adapter=SQLAdapter(dialect="postgresql"))

# Simple queries
result = nlp.transform("Pokaż wszystkich użytkowników")
print(result.command)  # SELECT * FROM users;

# Complex queries
result = nlp.transform("Pokaż użytkowników z Warszawy, którzy zarejestrowali się w ostatnim miesiącu")
print(result.command)  # SELECT * FROM users WHERE city = 'Warszawa' AND created_at >= NOW() - INTERVAL '1 month';

Advanced SQL (examples/sql/advanced_sql.py)

Joins and subqueries
Aggregation functions
Window functions
Complex filtering

Shell Examples

Basic Shell (examples/shell/basic_shell.py)

from nlp2cmd import NLP2CMD, ShellAdapter

nlp = NLP2CMD(adapter=ShellAdapter())

# File operations
result = nlp.transform("Znajdź pliki większe niż 100MB")
print(result.command)  # find . -size +100M -type f

# Process management
result = nlp.transform("Pokaż procesy zużywające najwięcej pamięci")
print(result.command)  # ps aux --sort=-%mem | head

Environment Analysis (examples/shell/environment_analysis.py)

System monitoring
Resource usage
Network analysis
Security checks

Docker Examples

Basic Docker (examples/docker/basic_docker.py)

from nlp2cmd import NLP2CMD, DockerAdapter

nlp = NLP2CMD(adapter=DockerAdapter())

# Container operations
result = nlp.transform("Pokaż wszystkie kontenery")
print(result.command)  # docker ps -a

# Image management
result = nlp.transform("Usuń nieużywane obrazy Docker")
print(result.command)  # docker image prune -f

Kubernetes Examples

Basic Kubernetes (examples/kubernetes/basic_kubernetes.py)

from nlp2cmd import NLP2CMD, KubernetesAdapter

nlp = NLP2CMD(adapter=KubernetesAdapter())

# Deployment operations
result = nlp.transform("Skaluj deployment nginx do 3 replik")
print(result.command)  # kubectl scale deployment nginx --replicas=3

# Status checking
result = nlp.transform("Pokaż wszystkie pody w namespace default")
print(result.command)  # kubectl get pods -n default

🏭 Real-World Use Cases

DevOps Automation

File: examples/use_cases/devops_automation.py

from nlp2cmd.generation import HybridThermodynamicGenerator

async def devops_workflow():
    generator = HybridThermodynamicGenerator()
    
    # System health check
    health = await generator.generate("Sprawdź status wszystkich usług")
    
    # Log analysis
    logs = await generator.generate("Znajdź błędy w logach aplikacji")
    
    # Resource optimization
    optimization = await generator.generate("Zoptymalizuj zużycie zasobów")
    
    return {
        "health": health,
        "logs": logs,
        "optimization": optimization
    }

Scenarios covered:

System monitoring
Log analysis
Resource optimization
Automated deployments
Backup procedures

Data Science & ML

File: examples/use_cases/data_science_ml.py

# Data analysis workflows
result = await generator.generate("Analizuj trendy sprzedaży z ostatniego kwartału")

# Model optimization
result = await generator.generate("Zoptymalizuj hiperparametry modelu")

# Data preprocessing
result = await generator.generate("Wyczyść i przygotuj dane do analizy")

Scenarios covered:

Data analysis
Model training
Feature engineering
Data validation
Pipeline automation

Healthcare Applications

File: examples/use_cases/healthcare.py

# Patient data analysis
result = await generator.generate("Analizuj dane pacjentów z grupy ryzyka")

# Treatment optimization
result = await generator.generate("Zoptymalizuj harmonogram leczenia")

# Resource allocation
result = await generator.generate("Rozdziel personel medyczny zgodnie z obciążeniem")

Scenarios covered:

Patient data analysis
Treatment scheduling
Resource optimization
Medical imaging
Clinical workflows

Finance & Trading

File: examples/use_cases/finance_trading.py

# Portfolio optimization
result = await generator.generate("Zoptymalizuj portfel inwestycyjny")

# Risk analysis
result = await generator.generate("Analizuj ryzyko kredytowe")

# Trading strategies
result = await generator.generate("Zaproponuj strategię handlową")

Scenarios covered:

Portfolio optimization
Risk assessment
Trading algorithms
Fraud detection
Compliance checking

Smart Cities

File: examples/use_cases/smart_cities.py

# Traffic optimization
result = await generator.generate("Zoptymalizaj sygnalizację świetlną")

# Energy management
result = await generator.generate("Zarządzaj zużyciem energii w mieście")

# Public transport
result = await generator.generate("Zaplanuj trasy transportu publicznego")

Scenarios covered:

Traffic management
Energy optimization
Public transport
Waste management
Urban planning

🌐 Browser Automation Examples (v1.0.85+)

API Key Extraction (Known Services — No LLM)

# Extract API key from OpenRouter and save to .env
nlp2cmd -r "otwórz openrouter.ai i wyciągnij klucz API, zapisz do .env"

# Same for Anthropic
nlp2cmd -r "wyciągnij klucz API z anthropic i zapisz do .env"

# GitHub token
nlp2cmd -r "pobierz token z github i zapisz do .env"

Supported services (rule-based, 0ms overhead): OpenRouter, Anthropic, OpenAI, GitHub, HuggingFace, Replicate.

# Open multiple tabs
nlp2cmd -r "otwórz 3 taby: github.com, gmail.com i stackoverflow.com"

Browser Automation with Video Recording

# Draw ladybug on jspaint with video recording
nlp2cmd -r "wejdź na jspaint.app i narysuj biedronkę" --video webm

# Record MP4
nlp2cmd -r "otwórz stronę example.com i wypełnij formularz" --video mp4

Multi-Step Browser Commands (Python API)

from nlp2cmd.generation.complex_detector import ComplexQueryDetector
from nlp2cmd.automation.action_planner import ActionPlanner

# Detect if query is multi-step
detector = ComplexQueryDetector()
result = detector.analyze(
    "otwórz przeglądarkę i stronę openrouter.ai, "
    "wyciągnij klucz API i zapisz do .env"
)
print(result.is_complex)    # True
print(result.num_intents)   # 4
print(result.intents)       # ['browser:launch', 'browser:navigate',
                            #  'browser:extract_data', 'browser:save_file']

# Decompose into action plan
planner = ActionPlanner()
plan = planner.decompose_sync(
    "otwórz openrouter.ai i wyciągnij klucz API, zapisz do .env"
)
for step in plan.steps:
    print(f"  {step.action}: {step.description}")
# navigate: Przejdź na stronę kluczy openrouter
# extract_api_key: Wyciągnij klucz API openrouter
# save_env: Zapisz OPENROUTER_API_KEY do .env

Execute Action Plan with PipelineRunner

from nlp2cmd.pipeline_runner import PipelineRunner

runner = PipelineRunner(headless=False, video_fmt="webm")
result = runner.execute_action_plan(plan, dry_run=False)
print(result.success)
print(result.data.get("video"))  # path to recorded video

Full Pipeline with Multi-Step Detection

from nlp2cmd.generation.pipeline import RuleBasedPipeline

pipeline = RuleBasedPipeline()
result = pipeline.process(
    "otwórz openrouter.ai i wyciągnij klucz API, zapisz do .env"
)

if result.action_plan:
    # Multi-step command detected
    print(f"Plan: {len(result.action_plan.steps)} steps")
    print(f"Source: {result.source}")  # "rule_decomposer" or "llm_planner"
else:
    # Single command
    print(f"Command: {result.command}")

🔬 Advanced Examples

Thermodynamic Optimization Benchmarks

File: examples/use_cases/physics_simulations.py

from nlp2cmd.generation import HybridThermodynamicGenerator

async def complex_optimization():
    generator = HybridThermodynamicGenerator()
    
    # Complex scheduling problem
    result = await generator.generate(
        "Zaplanuj 50 zadań w 20 slotach z ograniczeniami zasobów"
    )
    
    if result['source'] == 'thermodynamic':
        print(f"Energy: {result['result'].energy}")
        print(f"Samples: {result['result'].n_samples}")
        print(f"Solution: {result['result'].decoded_output}")

Multi-Step Workflows

File: examples/pipelines/log_analysis.py

from nlp2cmd import PlanExecutor, ExecutionPlan, PlanStep

executor = PlanExecutor()

# Multi-step log analysis pipeline
plan = ExecutionPlan(steps=[
    PlanStep(action="shell_find", params={"glob": "*.log"}, store_as="log_files"),
    PlanStep(action="shell_count_pattern", foreach="log_files", 
             params={"file": "$item", "pattern": "ERROR"}, store_as="error_counts"),
    PlanStep(action="summarize_results", params={"data": "$error_counts"}),
])

result = executor.execute(plan)

Configuration Management

File: examples/validation/config_validation.py

from nlp2cmd import SchemaRegistry

registry = SchemaRegistry()

# Validate configuration files
validation = registry.validate(content, "docker_compose")

# Repair configuration
repair = registry.repair(content, "docker_compose", auto_fix=True)

if repair["changes"]:
    print("Fixed issues:")
    for change in repair["changes"]:
        print(f"  - {change['reason']}")

🎯 Learning Path

Beginner Level

Start with CLI examples (shell_commands_demo.sh)
- Learn basic syntax
- Understand DSL types
- Try interactive mode
Conceptual overview (simple_demo_examples.py)
- Understand Python API concepts
- Learn shell command patterns
- Explore use cases

Intermediate Level

Python API basics (complete_python_shell_examples.py)
- HybridThermodynamicGenerator
- Context-aware queries
- Error handling
Domain-specific examples
- SQL: examples/sql/basic_sql.py
- Shell: examples/shell/basic_shell.py
- Docker: examples/docker/basic_docker.py

Advanced Level

Real-world use cases
- DevOps: examples/use_cases/devops_automation.py
- Data Science: examples/use_cases/data_science_ml.py
- Healthcare: examples/use_cases/healthcare.py
Complex workflows
- Thermodynamic: examples/use_cases/physics_simulations.py
- Pipelines: examples/pipelines/log_analysis.py
- Architecture: examples/architecture/end_to_end_demo.py

🛠️ Running Examples

Prerequisites

# Install NLP2CMD
pip install nlp2cmd

# For development mode
cd nlp2cmd
pip install -e .

CLI Examples

# Make shell demo executable
chmod +x examples/use_cases/shell_commands_demo.sh

# Run CLI demo
./examples/use_cases/shell_commands_demo.sh

# Individual CLI commands
nlp2cmd --query "show all users"
nlp2cmd --dsl docker --query "list containers"
nlp2cmd analyze-env

Python Examples

# Run Python examples
python3 examples/use_cases/simple_demo_examples.py
python3 examples/use_cases/complete_python_shell_examples.py
python3 examples/sql/basic_sql.py
python3 examples/use_cases/devops_automation.py

Development Mode

# Set PYTHONPATH for development
export PYTHONPATH=/path/to/nlp2cmd/src:$PYTHONPATH

# Run with development version
python3 examples/use_cases/complete_python_shell_examples.py

📊 Performance Benchmarks

DSL Generation

# Simple queries: ~2-5ms
await generator.generate("show users")

# Medium complexity: ~5-15ms
await generator.generate("find files larger than 100MB modified in last week")

# Complex queries: ~15-50ms
await generator.generate("analyze system logs for security threats")

Thermodynamic Optimization

# Simple optimization: ~100-200ms
await generator.generate("optimize 5 tasks in 10 slots")

# Medium complexity: ~200-500ms
await generator.generate("allocate resources to 20 projects")

# Complex optimization: ~500-2000ms
await generator.generate("optimize city traffic with 1000 intersections")

🔧 Customization

Adding Custom Examples

Create new example file

# examples/use_cases/my_custom_example.py
from nlp2cmd.generation import HybridThermodynamicGenerator

async def custom_workflow():
    generator = HybridThermodynamicGenerator()
    # Your custom logic here
    pass

Add to documentation
- Update this guide
- Add to README examples section
- Include in tests if needed

Extending DSL Support

# Custom adapter example
from nlp2cmd import BaseAdapter

class CustomAdapter(BaseAdapter):
    def transform(self, text: str, context: dict = None):
        # Custom transformation logic
        return "custom command"

🤝 Contributing Examples

When contributing examples:

Follow the structure - Use existing patterns
Add documentation - Explain the use case
Include tests - Unit tests for new examples
Update guides - Keep documentation current
Performance notes - Include timing information

📚 Additional Resources

CLI Reference - Complete CLI documentation
Python API Guide - Detailed API usage
User Guide - Complete tutorial
Thermodynamic Integration - Advanced optimization

🆘 Getting Help

Issues: GitHub Issues
Discussions: GitHub Discussions
Documentation: Full Docs
Examples: Browse the examples/ directory for more

This site is open source. Improve this page.

nlp2cmd

Examples Guide

📚 Related Documentation

🗂️ Examples Structure

🚀 Quick Start Examples

1. CLI Usage Examples

2. Python API Concepts

3. Complete Python API

4. Thermodynamic Computing Demo

📚 Domain-Specific Examples

SQL Examples

Shell Examples

Docker Examples

Kubernetes Examples

🏭 Real-World Use Cases

DevOps Automation

Data Science & ML

Healthcare Applications

Finance & Trading

Smart Cities

🌐 Browser Automation Examples (v1.0.85+)

API Key Extraction (Known Services — No LLM)

Multi-Tab Navigation

Browser Automation with Video Recording

Multi-Step Browser Commands (Python API)

Execute Action Plan with PipelineRunner

Full Pipeline with Multi-Step Detection

🔬 Advanced Examples

Thermodynamic Optimization Benchmarks

Multi-Step Workflows

Configuration Management

🎯 Learning Path

Beginner Level

Intermediate Level

Advanced Level

🛠️ Running Examples

Prerequisites

CLI Examples

Python Examples

Development Mode

📊 Performance Benchmarks

DSL Generation

Thermodynamic Optimization

🔧 Customization

Adding Custom Examples

Extending DSL Support

🤝 Contributing Examples

📚 Additional Resources

🆘 Getting Help