Video SEO title and chapter generator that outputs optimized titles and YouTube chapters
Views
4.3K
Copies
1.1K
Likes
388
Comments
0
Copy rate
26.0%
Prompt
You are a senior backend engineer at a creator-tools SaaS. Write a production-ready TypeScript module that generates YouTube-optimized video titles and chapter markers from a transcript.
The module must:
1. Export a class `VideoSeoGenerator` with typed public methods
2. Take a [transcript_format] transcript as input, along with the target [seo_goal]
3. Produce [title_variant_count] title variants scored by predicted CTR signals (curiosity, specificity, length)
4. Produce YouTube chapter markers with timestamps in `MM:SS` or `HH:MM:SS` format
5. Guarantee first chapter is `00:00` as required by YouTube
6. Merge chapter topics shorter than 30 seconds into the previous chapter
7. Validate input, return structured typed errors, never throw raw errors
8. Include JSDoc comments on all public methods
9. Be testable — inject a pure scoring function, no network calls
Use strict TypeScript. No `any`. Follow clean architecture. Output at least 80 lines of real code, not placeholders.Customise this prompt
Fill in 3 variables to personalise this prompt
Preview
You are a senior backend engineer at a creator-tools SaaS. Write a production-ready TypeScript module that generates YouTube-optimized video titles and chapter markers from a transcript.
The module must:
1. Export a class `VideoSeoGenerator` with typed public methods
2. Take a [transcript_format] transcript as input, along with the target [seo_goal]
3. Produce [title_variant_count] title variants scored by predicted CTR signals (curiosity, specificity, length)
4. Produce YouTube chapter markers with timestamps in `MM:SS` or `HH:MM:SS` format
5. Guarantee first chapter is `00:00` as required by YouTube
6. Merge chapter topics shorter than 30 seconds into the previous chapter
7. Validate input, return structured typed errors, never throw raw errors
8. Include JSDoc comments on all public methods
9. Be testable — inject a pure scoring function, no network calls
Use strict TypeScript. No `any`. Follow clean architecture. Output at least 80 lines of real code, not placeholders.