I am looking for an expert developer to create a stable automation tool for a desktop environment (running 10 emulator instances). The tool must use OpenCV for image recognition and behave like a human user. Key Requirements: Multi-Instance: Manage 10 windows simultaneously with optimized CPU/RAM usage. Core Logic: Detect specific icons on the map and perform automated clicks (5 concurrent actions per window). Security: Use randomized delays and curved mouse movements to mimic human behavior. Captcha Solver: Integrated system to handle visual puzzle verifications automatically. Remote Control (Discord): The tool must be controllable via Discord commands: !Start / !Stop: To control the operation. !View: To receive a real-time screenshot and status update on my phone. Emergency Response: Automated action if a specific visual alert is detected on screen. Deliverables: Clean, documented Python code and a setup guide for 24/7 stability.