Leaderboard Deploy agents

Distinguish between different categories of errors and handle each appropriately. Specifically:

  1. Validation errors occur during normal operation when input data doesn’t match expectations. These should be caught and handled gracefully.
  2. Usage errors indicate developer mistakes in using your API and generally shouldn’t be caught.

Document these distinctions clearly for API users and provide correct alternatives for common error scenarios.

# BAD: Catching usage errors that indicate developer mistakes
try:
    # Code that might produce both validation and usage errors
    result = model(invalid_input)
except Exception:  # Too broad!
    log_error("An error occurred")
    
# GOOD: Only catch expected validation errors
try:
    result = model(user_input)
except ValidationError as e:
    # Handle validation errors gracefully
    log_error(f"Validation failed: {e}")
    show_user_friendly_message()
    
# Usage errors (like PydanticUserError) aren't caught,
# allowing them to crash during development when they can be fixed

# In documentation, clearly explain error categories:
"""
## Error Types

### Validation Errors
Errors that happen during data validation when user input doesn't match schema.
These can be caught and handled at runtime.

### Usage Errors 
Errors that happen when using the API incorrectly.
These indicate developer mistakes and shouldn't be caught.
"""