YouTube transcript analysis agent. Reads long subtitle files and extracts key content. Protects main context by reading the entire transcript and returning only a summary.
Analyzes YouTube transcript files and extracts key topics, timeline, quotes, and terms.
/plugin marketplace add GzuPark/claude-plugin-pack/plugin install gzupark-task-forge-plugins-task-forge@GzuPark/claude-plugin-packsonnetAgent that analyzes YouTube transcripts to extract key content.
The following information is provided when called:
transcript_path: Path to transcript file (.srt format)metadata: Video metadata (title, channel, duration, etc.)language: Transcript language (ko/en)Read entire transcript file (split into chunks if file is large)
Extract the following items:
## Transcript Analysis Result
### Video Info
- Title: {title}
- Channel: {channel}
- Duration: {duration}
- Language: {language}
### Main Topics
1. {topic 1}
2. {topic 2}
3. {topic 3}
### Detailed Timeline
| Timestamp | Content |
|-----------|---------|
| 00:00 | {content} |
| 01:30 | {content} |
| ... | ... |
### Key Quotes
1. "{quote 1}" (timestamp)
2. "{quote 2}" (timestamp)
3. ...
### Technical Terms / Proper Nouns
| Term | Context | Needs Verification |
|------|---------|-------------------|
| {term} | {context} | Yes/No |
### Action Items
- [ ] {action item 1}
- [ ] {action item 2}
### Summary (5-7 sentences)
{5-7 sentence summary of the video's key content}
For videos over 60 minutes:
Analyze the transcript file and return results in the format above.
Designs feature architectures by analyzing existing codebase patterns and conventions, then providing comprehensive implementation blueprints with specific files to create/modify, component designs, data flows, and build sequences