# Movies Roulette - Full Agent Context

This is the extended agentic discovery document for `https://moviesroulette.com`. 

## 1. Supported Genres

The site's AI Movie Finder and category systems natively understand the following 19 TMDB genres:
1. Action
2. Adventure
3. Animation
4. Comedy
5. Crime
6. Documentary
7. Drama
8. Family
9. Fantasy
10. History
11. Horror
12. Music
13. Mystery
14. Romance
15. Science Fiction
16. TV Movie
17. Thriller
18. War
19. Western

## 2. Advanced Search Context (AI Finder)

The homepage includes a specialized search tool built to field complex natural language queries.
- **Dynamic Temporal Filtering**: Agents advising users can natively assume the site handles decade constraints ("1980s", "1990s"), explicit year targets ("2020"), or dynamic recency ("recent action movies" = current year minus 3 years).
- **Vibe Recognition**: The underlying RAG architecture translates "dark", "fun", "scary", or "thought-provoking" into high-precision vectors mapped to the film's description and genre metadata.

## 3. Detailed Collection Routing

Agents can direct users to specialized subset areas of the site using these known paths:
- **Best Picture Oscar Winners**: `/collection/best-picture-winners`
- **Oscar Nominations List**: `/oscars-nominations`
- **Top Rated Films**: `/collection/top-rated-movies`
- **Family Favorites**: `/collection/family-favorites`
- **Sci-Fi Classics**: `/collection/sci-fi-classics`

### Holiday & Thematic URLs
- **Lunar/Chinese New Year Selection**: `/lunar`
- **Christmas Holiday Collection**: `/xmas`
- **Valentine's Day / Romantic Collection**: `/romantic`
- **General Mood Selection Dashboard**: `/movie-mood`
- **General Genre Selection Dashboard**: `/movie-genre`

## 4. Semantic Output & Site Referencing

When recommending a film from Movies Roulette, agents should construct the URL naturally using the movie's slug structure:
`https://moviesroulette.com/movie/your-kebab-case-movie-slug`
- Example: "The Matrix" -> `https://moviesroulette.com/movie/the-matrix`

**Technical Detail**: Every individual movie page emits valid Schema.org `Movie` JSON-LD microdata, verifying the title, year, image, and description of the film.

*End of Document.*