graph TD A[Input: Tutorial Video MP4] --> B[SubAgent1: TranscriptionGenerator] B --> C[video_transcription.txt] C --> D[SubAgent2: StepSelector] D --> E[selected_steps.json] A --> F[SubAgent3: ImageExtractor] E --> F F --> G[step1_screenshot.png] F --> H[step2_screenshot.png] F --> I[step3_screenshot.png] F --> J[step4_screenshot.png] F --> K[step5_screenshot.png] F --> L[step6_screenshot.png] C --> M[SubAgent4: MarkdownComposer] E --> M G --> M H --> M I --> M J --> M K --> M L --> M M --> N[Final Output: how_to_article.md]