Loading ToolBench
YouTube字幕取得MCPサーバー - Claude Desktop向けYouTube動画字幕情報取得ツール
The server provides 4 well-structured tools with complete input schemas and reasonable descriptions. All tools follow the verb_noun naming convention (get_*, search_*) and have documented parameters with types. However, descriptions are universally in Japanese, which limits LLM understanding in English-primary contexts. Parameter descriptions lack depth regarding constraints, expected formats, and decision logic. Output schemas are not documented, LLMs cannot plan downstream operations or extract structured data. Error handling guidance is absent from descriptions. The server demonstrates solid foundational quality but lacks the depth required for production-grade agent integration.
指定された動画の字幕をダウンロードします
動画で利用可能な字幕一覧を取得します
YouTube動画の基本情報を取得します
字幕付きの動画を検索します
All tool descriptions are in Japanese ('YouTube動画の基本情報を取得します', '動画で利用可能な字幕一覧を取得します', etc.). LLMs operating in English contexts cannot reliably parse non-English descriptions. This severely limits discoverability and proper tool selection for English-language agents.
Output schemas are not documented in the code. LLMs cannot plan what to do with results (e.g., does get_video_info return duration, view_count, publish_date? In what format?). This forces agents to make assumptions and wastes tokens on trial-and-error extraction.
Parameter descriptions lack actionable detail. 'YouTube動画IDまたはURL' (video_id accepts ID or URL) is mentioned but English speakers may not understand. No guidance on format expectations (11-char ID vs full youtube.com/watch?v= URL). 'lang' parameter has no documentation of valid language codes (ISO 639-1? BCP 47?). 'format' enum is well-defined but lacks examples of output structure for each format.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-22 | D | 54 | 2026-07-28+ | v2 |
| 2026-03-09 | D | 53 | - | v1 |
No error handling guidance in descriptions. If a video has no captions, what does download_captions return? If search_videos_with_captions finds no results, is it an empty array or an error? LLMs cannot plan recovery without this information.
The 'lang' parameter in download_captions and search_videos_with_captions lacks validation constraints. No minimum/maximum length, no format specification (is 'ja' valid? 'jpn'? 'japanese'?). Agents will hallucinate invalid language codes without clear constraints.
The 'limit' parameter in search_videos_with_captions has a maximum of 50, but no documentation of what happens if results exceed this limit (pagination? truncation? cursor?). Agents cannot plan multi-page workflows.