Melting Face 🫠

A Python SDK that allows you to load prompts hosted on meltingface.eu. The library handles:

• Simple prompt objects (Prompt)
• Fetching public and private prompts from the MeltingFace hub
• Caching to avoid repeated downloads

Installation

pip install meltingface

Basic Usage

Loading a Public Prompt

from meltingface import Prompt

# Load a prompt from the public hub
prompt = Prompt.from_hub("owner/repo", version="0.1.0")

print(prompt.text)   # -> "Hello from my prompt!"
print(prompt.version) # -> "v1"

Loading a Private Prompt

Private prompts require authentication with an API key:

from meltingface import Prompt

# Option 1: Pass the API key directly
prompt = Prompt.from_hub("owner/private-repo", api_key="your-api-key")

# Option 2: Set the MELTINGFACE_API_KEY environment variable
# export MELTINGFACE_API_KEY="your-api-key"
prompt = Prompt.from_hub("owner/private-repo")

Coming Soon: API key support for accessing private prompts will be available on meltingface.eu as a paid feature. This will allow you to create, manage, and use private prompts programmatically in your applications.

Premium Subscription Benefits:

• Generate API keys to access your private prompts programmatically
• Use version-controlled prompts in programmatically
• Manage team access to shared prompts

Stay tuned for our official launch with multiple subscription tiers for individuals and teams.

API Reference

• repo_id ("owner/repo") is how you reference your prompt's repository on meltingface.eu
• version is optional; if not provided, the library defaults to "latest"
• api_key is required for private prompts; can be passed directly or via the MELTINGFACE_API_KEY environment variable
• cache_dir lets you specify a custom cache location (defaults to ~/.meltingface/prompts)
• force_download when set to True, bypasses the cache and fetches the prompt again