| This software has been designed to be an agent of code-ai.nvim |
This project is a monorepo built using NX, designed to facilitate interaction with various AI models, including ChatGPT and Gemini. The architecture is modular, allowing for easy scalability and maintainability, with shared libraries and utilities that streamline the development process. The project leverages TypeScript for type safety and clarity, and it incorporates various tools such as Jest for testing, ESLint for code quality, and SQLite for data storage, ensuring a robust and efficient development environment.
The ChatGPT agent, located in the ./chatgpt-agent/ directory, is designed to interact with OpenAI's ChatGPT model. It utilizes an Express server to handle HTTP requests and responses, allowing users to send messages and receive AI-generated responses. The agent is built around a repository pattern that manages the storage and retrieval of conversation data, ensuring that interactions are contextually aware. The implementation includes a dedicated HTTP client that abstracts the complexities of API communication, making it easier to integrate and extend the functionality of the ChatGPT agent.
The Gemini agent, found in the ./gemini-agent/ directory, serves a similar purpose but is tailored to interact with Google's Gemini model. Like the ChatGPT agent, it is built on an Express framework and follows a modular architecture that promotes code reuse and separation of concerns. The Gemini agent also employs a repository pattern for managing conversation history and settings, ensuring that user interactions are preserved across sessions. The agent's HTTP client is specifically configured to communicate with the Gemini API, providing a seamless experience for users looking to leverage the capabilities of this advanced AI model.
Requirements:
- Have Node.js installed on your machine.
- Have the API keys in
GEMINI_API_KEYandOPENAI_API_KEYenvironment variables.
Check ./code-ai-agent/code-ai-agent-monorepo for exemple for a launch script.
But manually, you can do the following:
cd code-ai-agent/
npm install nx@latest --save-dev
npm install @nx/node --save-devnpx nx reset
npx nx build lib
npx nx build gemini-agent
npx nx build chatgpt-agentnpx nx serve chatgpt-agent
npx nx serve gemini-agentOr, leverage Nx parallelism to build and serve all projects:
npx nx reset
npx nx run-many --target=build --projects=chatgpt-agent,gemini-agent,lib
npx nx run-many --target=serve --projects=chatgpt-agent,gemini-agent(This sections is for archives only)
npx create-nx-workspace@latest code-ai-agent --preset=apps
cd code-ai-agent/
npm install nx@latest --save-dev
npm install @nx/node --save-devnpx nx generate @nx/node:application gemini-agent
npx nx generate @nx/node:application chatgpt-agent
npx nx generate @nx/node:library lib