Skip to main content

GPT Plugins for 4all

Project description

GPT Plugins 4All GPTPlugins4All logo

GPT Plugins 4All is a Python library designed to facilitate the integration of GPT and other large language models with various APIs, leveraging OpenAPI specifications. This library simplifies the process of parsing OpenAPI specs, managing different authentication methods, and dynamically interacting with APIs based on model responses. PyPI version Demo using the AlphaVantage API with OpenAI

Features

  • Parse and validate OpenAPI 3.1.0 specifications.
  • Handle diverse authentication methods, including OAuth 2.0, Basic Auth, Header Auth, and Query Parameter Auth.
  • Generate structured API representations for AI interactions.
  • Dynamically construct API calls based on OpenAPI specs.
  • Support OAuth2.0 flow for token acquisition and usage.
  • Easily create and manage instances of AI assistants and threads for interactive sessions
  • Command-Line Interface (CLI) for convenient management of configurations and interactions.

Installation

Install GPT Plugins 4All using pip:

pip install GPTPlugins4All

Using the CLI

The GPT Plugins 4All CLI provides a convenient way to manage configurations and interact with your APIs from the command line. ###Common Commands Search for Configurations

gpt-plugins-4all search --query "your_search_query"

Fetch a Specific Configuration

gpt-plugins-4all get --id "config_id_or_name"

List Your Configurations

gpt-plugins-4all my-configs --api-key "your_api_key"

Submit a New Configuration

gpt-plugins-4all submit-config --url "config_url" --auth-type "auth_type" --visibility "visibility" --api-key "your_api_key"

Usage

The CLI supports various operations such as searching for configurations, retrieving specific configurations, listing user configurations, and submitting new configurations. You can use these commands directly from your terminal to interact with the GPT Plugins 4All library.

For detailed usage and available options for each command, use the --help flag with any command:

gptplugins4all [command] --help

Quick Start

Initializing with an OpenAPI Specification

We support initializing with an OpenAPI Spec in two ways. One way is to just give the name of the spec from https://gptplugins4all.com like this:

config = Config('alpha_vantage')

We also support directly making a config from an OpenAPI spec.

from GPTPlugins4All.config import Config

# Initialize the Config object with your OpenAPI spec
spec_string = """..."""  # Your OpenAPI spec as a string
config = Config(spec_string)

Adding Authentication Methods

Add Basic Authentication

config.add_auth_method("BASIC", {"key": "your_api_key"})

Add OAuth Configuration

config.add_auth_method("OAUTH", {
    "client_id": "your_client_id",
    "client_secret": "your_client_secret",
    "auth_url": "https://example.com/auth",
    "token_url": "https://example.com/token",
    "redirect_uri": "https://yourapp.com/oauth-callback",
    "scope": "read write"
})

Generating Simplified API Representations

simplified_api = config.generate_simplified_api_representation()
print(simplified_api)

Generate Object for use with OpenAI functions

tools = config.generate_tools_representation()

Using the Assistant Class

The Assistant class (for now) provides a simplified interface between your plugins and various OpenAI models via the Assistants API.

Initializing the Assistant

from assistant import Assistant

# Create an assistant instance
my_assistant = Assistant(config, "My Assistant", "Your instructions", "model_name")

Interacting with the assistant

# Getting a response from the assistant
response = my_assistant.get_assistant_response("Your query here")
print(response)

OAuth Flow

auth_url = config.start_oauth_flow()
# Redirect the user to auth_url...

tokens = config.handle_oauth_callback(code_from_redirect)

Making API Calls

response = config.make_api_call("/endpoint", "GET", {"param": "value"})

Oauth

url = config5.start_oauth_flow() #use this url to get code first
callback = config5.handle_oauth_callback(code)
#example
response = config5.make_api_call_by_path(path, "POST", params=your_params, user_token=callback, is_json=True)

Contributing

Contributions are welcome! Please check out the contributing guidelines.

License

GPT Plugins 4All is released under the MIT License.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

gptplugins4all-1.0.35.tar.gz (14.3 kB view details)

Uploaded Source

Built Distribution

GPTPlugins4All-1.0.35-py3-none-any.whl (15.6 kB view details)

Uploaded Python 3

File details

Details for the file gptplugins4all-1.0.35.tar.gz.

File metadata

  • Download URL: gptplugins4all-1.0.35.tar.gz
  • Upload date:
  • Size: 14.3 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/4.0.2 CPython/3.9.16

File hashes

Hashes for gptplugins4all-1.0.35.tar.gz
Algorithm Hash digest
SHA256 cc200c7d57cab8f1b652ac57df0011eed56cfb9f5bed04a82f2aff34530f4063
MD5 1c5c3dbd886b0fab7ae458410552e061
BLAKE2b-256 1b1a2805bc16336dc8c69a03b55a2dd2839abd989a832fc3659adee8139e743b

See more details on using hashes here.

File details

Details for the file GPTPlugins4All-1.0.35-py3-none-any.whl.

File metadata

File hashes

Hashes for GPTPlugins4All-1.0.35-py3-none-any.whl
Algorithm Hash digest
SHA256 9fdf98117ad62d858ba3eebd0e3e0928b412608daf7103348e81581d5b7870b0
MD5 a2fdee56cd1d2d9eb99f526f9b2ee636
BLAKE2b-256 c8478e1aca5e8e78b8c2e2fb6e281fc6621158397986bacd311ac17bef045754

See more details on using hashes here.

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page