caishi
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
No arguments | |||
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": false
} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| bilibili/videoB | Get Bilibili video metadata (title, author, duration, stats, description) by BV id, video URL or b23.tv short link. |
| douyin/detailA | Get a Douyin video/post in full: caption, engagement stats, author, no-watermark play address, image gallery (for 图文 posts). Accepts share short link, video URL, or numeric id. |
| web/readB | Read any web page: open the URL and return its main text as plain text (uses your logged-in Chrome session). |
| xiaohongshu/noteA | Read a Xiaohongshu (Little Red Book) note: title, author, full text, likes/collects/comments, tags. Pass a full note URL (with xsec_token) when possible. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 4 tools
Each tool targets a clearly distinct platform (Bilibili, Douyin, Xiaohongshu) or generic web, so there is no overlap in purpose. An agent can select the right tool purely from the source platform.
All names follow a consistent platform/resource pattern (bilibili/video, douyin/detail, web/read, xiaohongshu/note). The second segment mixes nouns (video, detail, note) with a verb (read), a minor deviation from a strict convention.
Four tools is slightly thin but reasonable for a focused content-fetching server covering three specific platforms plus generic web. Each tool clearly earns its place with no redundancy.
The surface covers read/retrieve operations well but offers no search, listing, or discovery operations, so agents must already know a URL or id. This is a notable gap for a content platform aggregator.