openai/plugins-quickstartArchived
Plugins Quickstart
This project is a starter kit and boilerplate for developing plugins that extend ChatGPT functionality via external APIs. It serves as a reference implementation for building tools that enable large language models to perform tasks through sequenced API calls and structured data mapping.
The repository provides a template for integrating REST APIs and a guide for securing communication between a language model and third-party service endpoints. It focuses on the setup of identity verification and handshake mechanisms between an agent and a private backend server.
The implementation covers the definition of external API integrations and the development of tooling for language model agents. These capabilities include the execution of API calls and the configuration of authentication to secure communication with third-party applications.
Features
- ChatGPT Plugin Development - Serves as a comprehensive starter kit for developing plugins that extend ChatGPT via external APIs.
- Boilerplates - Provides a boilerplate project for the rapid development and deployment of ChatGPT plugins.
- External Service Integrations - Provides the capability to convert natural language instructions into executable calls to external services.
- LLM Tool Definitions - Defines API schemas and tool sets that enable LLMs to perform complex real-world tasks.
- LLM Tool Orchestration - Implements a framework for orchestrating external API calls to allow LLMs to complete complex tasks.
- OpenAPI Specifications - Provides OpenAPI specifications that allow language models to discover and call available API functions.
- API Request Authentication - Implements identity verification mechanisms to secure communication between the agent and third-party applications.
- Token-Based Authentication - Secures API access using token-based authentication in request headers to verify authorized agents.
- External API Integrations - Integrates natural language interfaces with third-party REST services to retrieve data or trigger actions.
- Agent Handshake Configurations - Provides a guide for configuring secure handshakes and identity verification between agents and backend servers.
- Request Chaining - Enables chaining of multiple API requests by passing outputs from one call as inputs to the next.
- Interface-Based Request Routing - Directs AI-generated requests to specific external endpoints based on API interface definitions.
- JSON-RPC Implementations - Uses the JSON-RPC protocol for structured communication between the AI agent and the backend server.
- Provider API Authentications - Offers a reference implementation for authenticating requests from a language model to third-party service providers.
- Natural Language Parameter Mapping - Implements logic to translate natural language user prompts into structured arguments for external API calls.
- Prompt-to-API Mapping Templates - Provides a structural template for mapping natural language prompts to structured REST API requests.
- Stateless REST APIs - Employs a stateless REST architecture to handle AI agent requests as independent transactions.
- More to explore - Provides starter templates for AI plugin development.
