mcp-openvision

# Simplified OpenVision MCP Server Stories ## Background The current OpenVision MCP server implements several tools for image analysis using OpenRouter. We need to simplify it to focus on just one tool (`image_analysis`) and make it more customizable for AI agents. ## User Stories ### Story 1: Simplify MCP Server to One Tool **As a** developer **I want** the MCP server to be simplified to just one core tool **So that** it's easier to maintain and use **Acceptance Criteria:** - Remove unnecessary tools (extract_text_from_image, compare_images, detect_objects) - Keep only the core image_analysis functionality - Update documentation to reflect the simplified approach - Ensure the server is still MCP-compliant ### Story 2: Allow AI Agent to Customize Chat Completion Request **As an** AI agent **I want** to customize OpenRouter chat completion requests **So that** I can provide more specific instructions for image analysis **Acceptance Criteria:** - Allow customization of relevant parameters in the chat completion request - Ensure the model field is determined by environment variables or configuration - Make the messages field optional but automatically fill them when needed - Keep the API simple and easy to understand ### Story 3: Improve Configuration Management **As a** server administrator **I want** simple configuration options for the server **So that** I can easily set up and customize it **Acceptance Criteria:** - Support environment variables for configuration - Allow overriding defaults through configuration files - Document all configuration options clearly - Implement reasonable defaults ### Story 4: Improve Error Handling and User Experience **As a** user of the MCP server **I want** clear error messages and good documentation **So that** I can troubleshoot issues easily **Acceptance Criteria:** - Implement proper error handling for OpenRouter API errors - Return user-friendly error messages - Add example usage in documentation - Include troubleshooting tips ## Implementation Plan 1. **Simplification Phase:** - Remove unused tools from server.py - Consolidate functionality into one core tool - Update initialization and documentation 2. **Customization Phase:** - Extend the image_analysis tool to accept customization parameters - Implement automatic filling of required fields - Add validation for parameters 3. **Configuration Phase:** - Add support for reading configuration from environment variables - Implement sane defaults - Document configuration options 4. **Testing & Documentation Phase:** - Test the simplified implementation - Update README and documentation - Add examples of usage