Rollekino MCP server
OfficialClick on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@Rollekino MCP server search for Aki Kaurismäki films from the 2000s"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Rollekino MCP server
Read and write the film archive at rollekino.fi: about 3400 reviewed films, each with a personal 1-10 rating alongside IMDb and Metascore.
Runs on the same host as the site, served at https://www.rollekino.fi/mcp by nginx in front of the Python server on 8600 and its OAuth login on 8602. It is not WordPress and exposes only film tools.
Using this server with a paid AI service costs money. Tool definitions and results are billed as input tokens, and an agent can call tools repeatedly on its own. You are responsible for every charge, so set spending limits with your provider. The author accepts no liability for any costs. SeeDISCLAIMER.md.
Tools
Reading: search_films, get_film, list_films, get_stats, list_terms, list_queue, get_connection_status.
Writing: set_rating, write_review, create_film, trash_film, lookup_tmdb.
list_films filters by year, genre, director, actor and rating range, orders by rating, title or date, and paginates.
Related MCP server: TMDB MCP Server
The site side
The film data comes from rollekino/v1, added to the theme in inc/hooks/movies-api.php. Writes go through that namespace rather than /wp/v2/movie, because the theme registers a REST field named meta for the Vue archive which shadows core's meta object.
create_film with a tmdb_id enriches from TMDB and OMDB using the helpers in inc/movie-enrich.php, shared with quick-review.php, so a film created here renders identically to one posted from the quick review page.
Setup
uv venv && uv pip install -e .export ROLLEKINO_URL=https://www.rollekino.fi
export ROLLEKINO_USER=... # WordPress user with edit_posts
export ROLLEKINO_APP_PASSWORD=... # application passwordReads are public; only writes need the credentials.
Available Tools
14 toolscreate_filmAIdempotent
Add a film to the archive, enriched from TMDB.
With tmdb_id the site fetches plot, poster, backdrop, cast, crew, genres, trailer, IMDb and Metascore exactly as the quick review page does. Find the id with lookup_tmdb.
The archive keeps one entry per film. If this one is already there, nothing is created and the existing entry comes back instead: update that with update_film or write_review rather than adding a second.
Args: title: Film title, usually the Finnish release title if it has one. tmdb_id: TMDB id, from lookup_tmdb. Strongly preferred. rating: Rolle's own rating: whole number 1 to 10, 0 for not rated. This is the archive's native scale, so never convert from IMDb or Metascore. If he gives stars out of 5, double them. date: Date shown on the review, YYYY-MM-DD or YYYY-MM-DD HH:MM, Helsinki time. Use the day he watched it, which Trakt history records, not today. content: His review text. Leave empty rather than writing one for him. status: draft or publish. imdb_url: Full IMDb URL, when there is no tmdb_id. year: Release year, when there is no tmdb_id. plot: Finnish synopsis, when there is no tmdb_id. allow_duplicate: Create even if the film exists. Almost never right.
Inputs: title (required), date, plot, year, +6 more. Returns: result
find_filmARead-onlyIdempotent
Find a film by its IMDb id, including drafts.
The reliable way to check whether a film is already in the archive before adding it, since titles repeat across remakes and translations.
Args: imdb_id: IMDb id such as tt0816692, or a full IMDb URL.
Inputs: imdb_id (required). Returns: result
get_connection_statusARead-onlyIdempotent
Report which site this points at and whether writing is possible.
Inputs: no named parameters. Returns: result
get_filmARead-onlyIdempotent
Get one film in full: the review, the rating, credits and every field.
Args: id: Film id, from search_films or list_films.
Inputs: id (required). Returns: result
get_statsARead-onlyIdempotent
Get archive-wide figures: totals, average rating, the rating spread and films per year.
Use this for questions about the collection rather than one film.
Inputs: no named parameters. Returns: result
list_filmsARead-onlyIdempotent
List films, filtered and ordered.
Args: year: Release year, as the archive records it. genre: Genre name in Finnish, such as Komedia or Draama. director: Director's name. actor: Actor's name. min_rating: Only films Rolle rated at least this, 1 to 10. max_rating: Only films he rated at most this, 1 to 10. unrated: True for films with no rating yet. imdb: IMDb id, to find one specific film. status: publish, draft, or any. orderby: date, title or rating. order: ASC or DESC. page: Which page of results. per_page: Results per page, at most 100.
Inputs: imdb, page, year, actor, +9 more. Returns: result
list_queueARead-onlyIdempotent
List draft films waiting for a review.
These are queued by the Trakt importer after Rolle watches something. Needs an application password.
Inputs: no named parameters. Returns: result
list_termsARead-onlyIdempotent
List genres, directors, writers or actors used in the archive.
Args: taxonomy: One of genre, director, writer, actor. search: Narrow to names containing this. limit: How many terms to return.
Inputs: taxonomy (required), limit, search. Returns: result
lookup_tmdbARead-onlyIdempotent
Search TMDB for a film, to get the tmdb_id that create_film enriches from.
This goes through the site, which holds the TMDB key, so no key is needed here. Needs an application password.
Args: query: Film title to search for.
Inputs: query (required). Returns: result
search_filmsARead-onlyIdempotent
Find films by title.
Start here when you know the name. Returns the id needed by get_film.
Args: title: Part of the film's title. limit: How many matches to return.
Inputs: title (required), limit. Returns: result
set_ratingBIdempotent
Set Rolle's own rating for a film.
Args: id: Film id. rating: Whole number 1 to 10, or 0 to clear. See create_film for the scale.
Inputs: id (required), rating (required). Returns: result
trash_filmADestructiveIdempotent
Move a film to the trash. It can be restored from the WordPress admin.
Args: id: Film id.
Inputs: id (required). Returns: result
update_filmAIdempotent
Change anything about an existing film. Only the fields given change.
Args: id: Film id. title: New title. date: Date shown on the review. See create_film. rating: Whole number 1 to 10, 0 for not rated. See create_film. status: publish or draft. Publishing a draft keeps its date. tmdb_id: Refresh every metadata field, poster and credits from TMDB. The rating and review text are kept. plot: Finnish synopsis. imdb_url: Full IMDb URL. genre: Replace the genres. Finnish names, as list_terms shows them. director: Replace the directors. writer: Replace the writers. actor: Replace the cast.
Inputs: id (required), date, plot, actor, +8 more. Returns: result
write_reviewAIdempotent
Put Rolle's review text on a film.
The text is his: pass what he wrote, never compose one for him.
Args: id: Film id. content: The review, as HTML or plain paragraphs. Replaces what is there. rating: Whole number 1 to 10. See create_film. date: Date shown on the review. See create_film. publish: True moves a draft to published.
Inputs: id (required), content (required), date, rating, +1 more. Returns: result
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 14 tools
Most tools have clearly distinct purposes keyed to a specific input (id vs IMDb id vs title) and action. There is mild overlap between list_films (which has an imdb filter and free-form filtering) and both search_films and find_film, but the descriptions do a good job steering the agent to the right one.
Every tool follows a consistent verb_noun pattern (get_film, list_films, create_film, update_film, trash_film, write_review, set_rating, find_film, lookup_tmdb). No mixed casing or verb styles.
14 tools is well within the sweet spot and each maps to a real operation: reads, CRUD, rating/review, taxonomy listing, stats, queue, TMDB lookup, and a health check. Nothing feels padded.
Covers full lifecycle: create/update/trash, get/list/search/find, rating, review, terms, stats, queue, and TMDB enrichment, plus a pre-add duplicate check. The only gap is no in-MCP restore from trash (only via the WordPress admin) and no explicit untrash tool.
Maintenance
Related MCP Connectors
Your film diary: log and rate films, search what you've watched, keep a watchlist and see stats.
Your film diary: log and rate films, search what you've watched, keep a watchlist and see stats.
Movie and series ranking as MCP tools: top titles, search, lookup by IMDb id. Free, no key.
OMDb MCP — IMDB-derived movie / TV / episode data (BYO key)
Related MCP Servers
- FlicenseNot gradedqualityDmaintenanceProvides movie listings, showtimes, and personalized recommendations for Amsterdam cinemas by scraping filmladder.nl, with support for filtering by date, cinema, rating, and preferred showtimes.1-
- FlicenseNot gradedqualityNot gradedmaintenanceProvides access to The Movie Database (TMDB) API for searching movies, retrieving movie details, cast information, and browsing current, upcoming, popular, and top-rated films, as well as searching for actors and directors.61 npm-
- AlicenseNot gradedqualityBmaintenanceEnables MCP clients to query a local movie library and IMDb catalog using shared filters, and to inspect stats or update, add, and remove library entries. It supports natural-language questions about owned, watched, rated, and unwatched films.9 npmMIT
- AlicenseNot gradedqualityCmaintenanceEnables movie and TV lookups through twelve read-only tools: title and person search, full details, episode lists, discovery by genre/year/language/streaming service, trending and recommendations, where to watch by country, and merged critic/audience scores from IMDb, Rotten Tomatoes, Metacritic, Letterboxd and others. Runs as a local stdio server or a containerized Streamable HTTP service with bearer auth.1MIT