graph TD
A[Input: MP4 Video URL] --> B[Subagent 1: VideoTranscriber]
B --> C[Generate Timestamped Transcription]
C --> D[Send to StepExtractor]
D --> E[Subagent 2: StepExtractor]
E --> F[Parse Transcription]
F --> G[Identify Key Actions]
G --> H[Condense to 6 Steps]
H --> I[Create Step/Timestamp Pairs]
I --> J[Subagent 3: VideoFrameGrabber]
J --> K[Extract Frame at Each Timestamp]
K --> L[Save as PNG/JPG]
L --> M[Name Files Systematically]
I --> N[Subagent 4: TextRefiner]
N --> O[Polish Step Descriptions]
O --> P[Generate Intro/Conclusion]
M --> Q[Subagent 5: ArticleAssembler]
P --> Q
Q --> R[Combine Text and Images]
R --> S[Format in HTML/Markdown]
S --> T[Output Final Article]