Skip to content

EXAMPLE.md

Latest commit

 

History

History
88 lines (65 loc) · 2.45 KB

EXAMPLE.md

File metadata and controls

88 lines (65 loc) · 2.45 KB

Esempi di Utilizzo

Analisi Prompt

from src.analyzers.token_analyzer import TokenAnalyzer
from src.analyzers.prompt_analyzer import PromptAnalyzer

# Analisi token
analyzer = TokenAnalyzer("gpt")
analysis = analyzer.analyze_toon_file("path/to/prompt.toon.md")
print(f"Token totali: {analysis['total_file_tokens']}")

# Analisi best practices
prompt_analyzer = PromptAnalyzer()
result = prompt_analyzer.analyze("path/to/prompt.toon.md")
print(f"Score best practices: {result['best_practices_score']}")

Ottimizzazione

from src.optimizers.token_optimizer import TokenOptimizer
from src.optimizers.structure_optimizer import StructureOptimizer

# Ottimizza token
token_opt = TokenOptimizer("gpt")
result = token_opt.optimize(prompt_text, target_reduction=0.20)
print(f"Riduzione: {result['reduction_percent']}%")

# Ottimizza struttura
struct_opt = StructureOptimizer()
result = struct_opt.optimize("path/to/prompt.toon.md", "path/to/optimized.toon.md")

Validazione Coerenza

from src.analyzers.consistency_checker import ConsistencyChecker

checker = ConsistencyChecker("path/to/prompt_engineering")
result = checker.check_consistency("senior-phd-devops-engineer")
print(f"Score coerenza: {result['consistency_score']}")
print(f"Competenze core: {result['core_competencies']}")

Generazione Test

from src.validators.test_generator import TestGenerator

generator = TestGenerator()
test_data = generator.generate("path/to/prompt.toon.md", "path/to/test.json")
print(f"Scenari generati: {len(test_data['scenarios'])}")

Import README

from src.analyzers.readme_importer import ReadmeImporter

# Importa tutto il contesto
importer = ReadmeImporter()
context = importer.import_to_context()

# Oppure importa selettivamente
readme_content = importer.read_readme()
platforms = importer.extract_platforms()
best_practices = importer.extract_best_practices()
toon_info = importer.extract_toon_format_info()

print(f"Piattaforme trovate: {len(platforms)}")
print(f"Best practices: {len(best_practices)}")

Utilizzo via MCP

Il tool è esposto come server MCP. Per utilizzarlo:

  1. Configura il server MCP nel tuo client
  2. Chiama i tools disponibili:
    • analyze_prompt: Analisi completa
    • optimize_prompt: Ottimizzazione
    • validate_consistency: Validazione cross-platform
    • token_analysis: Analisi dettagliata token
    • generate_tests: Generazione test
    • compare_prompts: Confronto due prompt