double-o 0.1.3

A Python library for secret management and API proxy calls

Double-O

A Python library for secret management and API proxy calls.

Installation

pip install double-o

Quick Start

Fetching Secrets

import oo

# Simple one-liner to fetch a secret
secret = oo.get_secret("YOUR_TOKEN_HERE")
print(f"Secret: {secret}")

Making Proxy API Calls

import oo

# Make an API call through the proxy
result = oo.proxy(
    "v1/chat/completions",
    token="YOUR_PROXY_TOKEN",
    payload={
        "model": "gpt-4o-mini",
        "messages": [{"role": "user", "content": "Hello!"}]
    }
)
print(result)

Chat Completions (Convenience Method)

import oo

# Even simpler for chat completions
result = oo.chat(
    token="YOUR_PROXY_TOKEN",
    messages=[{"role": "user", "content": "Hello!"}],
    model="gpt-4o-mini"
)
print(result)

Advanced Usage

Using the Client Class

For more control, use the Client class directly:

from oo import Client

# Create a client with custom settings
client = Client(
    base_url="http://localhost:3001",
    timeout=60
)

# Fetch a secret
secret = client.get_secret("YOUR_TOKEN")

# Make proxy calls
result = client.proxy(
    "v1/chat/completions",
    token="YOUR_PROXY_TOKEN",
    payload={"model": "gpt-4o-mini", "messages": []}
)

# Don't forget to close when done
client.close()

Context Manager

The client supports context managers for automatic cleanup:

from oo import Client

with Client(base_url="http://localhost:3001") as client:
    secret = client.get_secret("YOUR_TOKEN")
    # Client is automatically closed when exiting the block

Custom Base URL

All functions accept a base_url parameter:

import oo

# Use a different server
secret = oo.get_secret(
    "YOUR_TOKEN",
    base_url="http://your-server:8080"
)

Error Handling

The library provides custom exceptions for different error scenarios:

import oo
from oo import SecretError, ProxyError, AuthenticationError

try:
    secret = oo.get_secret("invalid_token")
except AuthenticationError as e:
    print(f"Authentication failed: {e}")
except SecretError as e:
    print(f"Failed to fetch secret: {e}")

try:
    result = oo.proxy("v1/endpoint", "invalid_token", payload={})
except AuthenticationError as e:
    print(f"Proxy authentication failed: {e}")
except ProxyError as e:
    print(f"Proxy request failed: {e}")

Exception Hierarchy

• DoubleOError - Base exception for all Double-O errors
  • SecretError - Raised when fetching a secret fails
  • ProxyError - Raised when a proxy request fails
  • AuthenticationError - Raised when authentication fails

API Reference

Functions

oo.get_secret(token, base_url="http://localhost:3001")

Fetch a secret value using a token.

• token (str): The authentication token
• base_url (str): API server URL (default: http://localhost:3001)
• Returns: The secret value as a string
• Raises: SecretError, AuthenticationError

oo.proxy(path, token, method="POST", payload=None, headers=None, base_url="http://localhost:3001")

Make an API call through the proxy.

• path (str): API path (e.g., 'v1/chat/completions')
• token (str): Proxy authentication token
• method (str): HTTP method (default: POST)
• payload (dict): Request payload (optional)
• headers (dict): Additional headers (optional)
• base_url (str): API server URL (default: http://localhost:3001)
• Returns: JSON response as a dictionary
• Raises: ProxyError, AuthenticationError

oo.chat(token, messages, model="gpt-4o-mini", base_url="http://localhost:3001", **kwargs)

Convenience function for OpenAI chat completions.

• token (str): Proxy authentication token
• messages (list): List of message dicts with 'role' and 'content'
• model (str): Model to use (default: gpt-4o-mini)
• base_url (str): API server URL (default: http://localhost:3001)
• kwargs: Additional parameters for the API
• Returns: Chat completion response

Client Class

Client(base_url="http://localhost:3001", timeout=30)

Create a new Double-O client.

• base_url (str): API server URL
• timeout (int): Request timeout in seconds

Methods:

• get_secret(token) - Fetch a secret
• proxy(path, token, method="POST", payload=None, headers=None) - Make proxy call
• chat_completion(token, messages, model="gpt-4o-mini", **kwargs) - Chat completion
• close() - Close the client session

Development

Setup

# Clone the repository
git clone https://github.com/yourusername/double-o.git
cd double-o

# Install development dependencies
pip install -e ".[dev]"

Running Tests

pytest

Code Formatting

black oo tests
isort oo tests

Type Checking

mypy oo

License

MIT License - see LICENSE for details.

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.