MCP Server for parsing Douyin share links — extract watermark-free video, cover, and music URLs
Two tools with basic schemas and Chinese descriptions. parse_douyin has a clear single parameter (share_url) with type and description; get_tiktok_trending has limit with type, default, and description. However, both descriptions lack English translations and are moderately detailed (100-150 chars). Output schemas are NOT documented, the code shows format_video_info() returns formatted text, but no structured schema is declared. No error handling guidance in descriptions. No parameter validation constraints (e.g., limit bounds for get_tiktok_trending). Tool names follow verb_noun pattern (parse_*, get_*), which is good. Risk annotations (READ_ONLY) are present but not exposed as toolAnnotations in the MCP schema.
获取 TikTok 热门视频列表,返回当前热门视频的标题、作者、播放量和点赞数。
解析抖音分享链接,提取视频信息。支持 v.douyin.com 短链接和 iesdouyin.com 分享链接。返回视频标题、作者、时长、封面URL、无水印下载URL、音乐URL。
Output schemas not documented. Code shows format_video_info() returns formatted text, but no structured response schema is declared in tool definitions. LLMs cannot plan downstream operations or extract fields reliably.
Descriptions are in Chinese only and lack English translations. LLMs trained primarily on English may misinterpret intent. Descriptions should be 50-200 chars and bilingual or English-primary.
No parameter constraints documented. get_tiktok_trending accepts 'limit' with default 15, but no min/max bounds are specified. LLMs may pass absurd values (0, 10000) that break the API.
Error handling descriptions missing. parse_douyin can fail (invalid URL, API changes, network errors) but tool description does not guide LLM on recovery steps or error categories.
Inferred effective spec: <=2025-11-25.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 49 | <=2025-11-25 | v2 |
Tool annotations (readOnlyHint, idempotentHint) not exposed in MCP schema. Risk field is present in source but not registered as toolAnnotations capability.