Autodocument MCP Server

# OpenRouter Client Documentation ## Overview The OpenRouter Client module provides a client for communicating with the OpenRouter API using the OpenAI SDK. It allows for generating documentation and content based on code files using large language models (LLMs). ## Key Components ### DocumentationResponse Interface This interface represents the response from the LLM after generating documentation: - `content`: The generated documentation content - `successful`: Boolean indicating if the generation was successful - `error`: Optional error message if generation failed ### OpenRouterClient Class This class handles communication with the OpenRouter API and provides methods for generating documentation. #### Constructor Creates a new OpenRouter client instance: - Accepts optional `apiKey` and `model` parameters that override configuration - Validates the API key's presence - Initializes an OpenAI client with OpenRouter-specific settings #### Key Methods 1. **generateWithCustomPrompt** - Generates content using a custom system prompt - Parameters: - `files`: Array of file objects with path and content - `systemPrompt`: Custom system prompt to use - `existingContent`: Optional existing content to update - `isTopLevel`: Whether this is the top level directory - `childrenContent`: Optional content from child directories - Returns a `DocumentationResponse` 2. **generateDocumentation** - Backward compatibility method that generates documentation for files - Automatically selects an appropriate system prompt based on context - Parameters similar to `generateWithCustomPrompt` - Internally calls `generateWithCustomPrompt` with the constructed prompt ## Implementation Details The client: - Uses environment variables or passed parameters for configuration - Sets required headers for OpenRouter API - Constructs appropriate prompts based on the directory level - Handles errors and returns structured responses - Adapts the prompt based on whether files exist in the current directory - Incorporates information from subdirectories when available - Considers existing documentation for updates The implementation ensures proper error handling and provides meaningful feedback when documentation generation fails.