nova-meta-prompter 0.1.0

Transform prompts to align with Amazon Nova guidelines

nova-meta-prompter

Transform prompts to align with Amazon Nova guidelines.

Installation

pip install nova-meta-prompter

Or install from source:

git clone <repository-url>
cd nova_metaprompter
pip install -e .

Quick Start

from nova_meta_prompter import transform_prompt

# Transform a prompt
result = transform_prompt("Summarize this document: {document_text}")

# Get the Nova-aligned prompt
print(result['nova_final'])

# View the transformation process
print(result['thinking'])      # Analysis of changes
print(result['nova_draft'])    # Initial draft
print(result['reflection'])    # Reflection on draft

Advanced Usage

Custom Model

# Use Claude Sonnet instead of default Nova Premier
result = transform_prompt(
    "Your prompt here",
    model_id='global.anthropic.claude-sonnet-4-5-20250929-v1:0'
)

Custom Boto Client

import boto3
from botocore.config import Config

# Create custom client
config = Config(region_name='us-east-1', read_timeout=2000)
client = boto3.client('bedrock-runtime', config=config)

# Use with transform_prompt
result = transform_prompt("Your prompt", boto_client=client)

Interactive Notebook

Explore the package with the included Jupyter notebook:

export AWS_PROFILE=your-profile
jupyter notebook nova_metaprompter_showcase.ipynb

AWS Configuration

Set your AWS credentials:

export AWS_PROFILE=your-profile-name
export AWS_REGION=us-west-2

Or use AWS CLI:

aws configure

Requirements

• Python 3.8+
• AWS Bedrock access with Nova model permissions
• Valid AWS credentials

What It Does

The transformer:

• Analyzes your existing prompts
• Identifies alignment opportunities with Nova guidelines
• Applies Nova best practices (clear headers, structured format, step-by-step instructions)
• Provides detailed reasoning and reflection on changes

Development

Run tests:

pip install -e ".[dev]"
pytest

License

MIT