Develop

[BYLinMou]

This pull request refactors and streamlines the API logic in server/src/routes/api.js, centralizing the AI call logic, improving configuration management, and enhancing error handling. It also updates the workflow configuration to only build Docker images on the main branch and bumps the app version. The most significant changes are grouped below:

API Refactoring and Centralization:

• Extracted and centralized the OpenAI API call logic into a new callOpenAI helper function, reducing code duplication and standardizing how prompts are sent and responses are parsed across endpoints. (server/src/routes/api.js) [1] [2]
• Added a getLanguageInstruction utility to consistently handle language-specific instructions for both standard and roleplay AI responses. (server/src/routes/api.js)
• Improved error handling and response structure for all AI endpoints, providing more consistent and informative error messages. (server/src/routes/api.js) [1] [2] [3]

Configuration Improvements:

• Enhanced the loadApiConfig function to require and validate a DEFAULT_MODEL configuration, ensuring the AI model is always specified and preventing misconfiguration. (server/src/routes/api.js)
• Cleaned up and clarified how models are selected for AI requests, always defaulting to the configured model and removing scattered fallback logic. (server/src/routes/api.js)

Workflow and Versioning:

• Updated the GitHub Actions workflow to only build and push Docker images on the main branch, not on develop. (.github/workflows/build.yml)
• Bumped the application version to 0.0.4 and updated the publish date in package.json. (package.json)

These changes make the backend codebase more maintainable, robust, and easier to extend for future AI-powered features.