BetaSeries MCP Server
An MCP server that exposes the full BetaSeries API as 23 tools, enabling LLM clients to track TV shows and movies, manage planning, ratings, and social features via natural language.
README
BetaSeries MCP Server
An open-source Model Context Protocol server for the BetaSeries API — TV show and movie tracking, planning, ratings, social features, and more. Self-hostable via Docker, or run locally over stdio with any MCP client.
Features
- 186 BetaSeries operations, grouped into 23 tools (one per API resource category) so an LLM client doesn't choke on tool count — every real endpoint the API exposes is reachable, with none of the noise of one-tool-per-endpoint.
- Generated straight from BetaSeries' own OpenAPI spec (
scripts/generate-tools.ts), so keeping up with API changes is a re-vendor +npm run codegen, not hand-editing 190 parameter lists. - Both stdio (local MCP clients) and Streamable HTTP (Docker / remote) transports from the same server.
- Auth is a one-time
npm run auth(OAuth2 device-code flow) — the running server never exposes a "log in" tool to the LLM.
Tool categories
| Tool | Operations | Covers |
|---|---|---|
betaseries_shows |
34 | search, display, tracking, ratings, episodes, seasons, similars, recommendations, videos, tags... |
betaseries_members |
24 | profile, options, notifications, badges, search, signup, sync, locale, avatar/banner... |
betaseries_movies |
21 | search, display, favorites, ratings, discover, upcoming, similars, characters... |
betaseries_episodes |
17 | list, watched, downloaded, hidden, rewatch, notes, next/latest, unrated... |
betaseries_comments |
11 | list, post, delete, replies, reactions, subscriptions... |
betaseries_pictures |
9 | poster/image URLs for shows, movies, episodes, seasons, members, characters, persons, platforms, badges |
betaseries_collections |
8 | manage & subscribe to collections (Premium) |
betaseries_seasons |
8 | watched, hidden, rewatch, notes |
betaseries_quiz |
7 | list, answer, history, rankings |
betaseries_friends |
6 | list, requests, add/remove, block |
betaseries_timeline |
6 | home/friends/member/show feeds, events |
betaseries_messages |
5 | inbox, discussion, send, read |
betaseries_polls |
5 | list, latest, answer, target |
betaseries_subtitles |
5 | episode/season/show subtitles, latest, report |
betaseries_platforms |
4 | list & manage streaming platform subscriptions |
betaseries_planning |
3 | your schedule, general & upcoming planning |
betaseries_search |
3 | unified search across shows/movies |
betaseries_tags |
3 | list & tag shows |
betaseries_persons |
2 | actor/director/crew info & articles |
betaseries_reports |
2 | report content, update a report |
betaseries_badges |
1 | badge details |
betaseries_news |
1 | latest news |
betaseries_stats |
1 | show ranking stats |
Each tool takes operation (an enum of that category's operation IDs) and params (a
string key/value map for that operation) — the full parameter reference for every
operation is in the tool's own MCP description, generated from the spec.
⚠️ A note on openapi.json
BetaSeries' published spec (https://developers.betaseries.com/openapi.json) contains 7
extra operations tagged "Model" — /discover, /resources, /resources/{id},
/query, /execute, /version, /setCredentials — worded like Model Context Protocol
primitives ("Discover MCP server capabilities", "Execute an action through the MCP
server", "Set API key... for LLM authentication"). They do not exist on the real API
(api.betaseries.com returns a plain Page not found. for all of them, unlike real
endpoints), and there's no BetaSeries announcement of an official MCP server. They look
like either leftover boilerplate from a spec-generation tool, or bait aimed at AI agents
auto-generating MCP servers from this exact file. Either way, this project strips them —
see spec/PROVENANCE.md for the full writeup. If BetaSeries ships
a real MCP server later, treat this project as a candidate for archival in favor of theirs.
Setup
1. Get a BetaSeries API key
Register an application at BetaSeries to get an API key (and secret, needed for step 3) — see developers.betaseries.com.
2. Install
git clone https://github.com/framinosona/betaseries-mcp.git
cd betaseries-mcp
npm install
cp .env.example .env
# edit .env: set BETASERIES_API_KEY (and BETASERIES_API_SECRET for the next step)
3. Authenticate (one-time)
npm run auth
This runs BetaSeries' OAuth2 device-code flow: it prints a code and a URL, you approve
it in your browser, and the resulting access token is saved into .env automatically as
BETASERIES_ACCESS_TOKEN.
4. Run
Locally over stdio (default):
npm run build
npm start
Or for development, without a build step:
npm run dev
5. Point an MCP client at it
For Claude Code / Claude Desktop, add to your MCP config:
{
"mcpServers": {
"betaseries": {
"command": "node",
"args": ["/path/to/betaseries-mcp/dist/index.js"],
"env": {
"BETASERIES_API_KEY": "...",
"BETASERIES_ACCESS_TOKEN": "..."
}
}
}
}
Self-hosting with Docker
cp .env.example .env # fill in your key/token as above
docker compose up -d
This runs the server in Streamable HTTP mode on http://localhost:3000/mcp
(/health for a liveness check). Point any MCP client that supports Streamable HTTP at
that URL. To run npm run auth against the containerized build instead of locally, use
docker compose run --rm betaseries-mcp npm run auth.
Environment variables
| Variable | Required | Default | Notes |
|---|---|---|---|
BETASERIES_API_KEY |
yes | — | from your BetaSeries API key |
BETASERIES_API_SECRET |
only for npm run auth |
— | from your BetaSeries API key |
BETASERIES_ACCESS_TOKEN |
for member-scoped operations | — | written by npm run auth |
BETASERIES_LOCALE |
no | fr |
matches the API's own default |
BETASERIES_API_BASE_URL |
no | https://api.betaseries.com |
|
BETASERIES_API_VERSION |
no | 3.0 |
|
MCP_TRANSPORT |
no | stdio |
stdio or http |
PORT |
no | 3000 |
HTTP mode only |
Development
npm run codegen # spec/betaseries-openapi.json -> src/generated/operations.ts
npm run typecheck
npm test
To pick up a BetaSeries API change: re-fetch https://developers.betaseries.com/openapi.json,
strip any "Model"-tagged operations (see spec/PROVENANCE.md), overwrite
spec/betaseries-openapi.json, then npm run codegen.
License
MIT — see LICENSE.md.
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。