captcha-bypass

This skill provides capabilities for solving CAPTCHA challenges including OCR text/math captchas, slide captchas, rotate captchas, and object detection captchas. Supports both CLI mode (direct command-line invocation) and Server mode (HTTP API + MCP endpoint for AI agent integration). It should be used when the task involves recognizing, bypassing, or solving any type of CAPTCHA — including text-based verification codes, arithmetic captchas, slider puzzles, rotate-to-align challenges, and YOLO-based object detection. The skill runs all inference on-device using ONNX models and OpenCV, no external API calls required.