Provides tools for retrieving Bilibili video details and comments, including real BVID details and comments.
Click on "Install 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., "@AI Social Data MCP What's trending on Douyin right now?"
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.
Built for Cursor · Claude · Codex
搜索、自动翻页、读取详情、分析作者与评论区——全部通过 MCP 完成。
一键安装到 Cursor · 在线服务 · npm · Release
普通网页搜索通常只能给 Agent 一组链接和摘要。面对社交平台，真正有价值的信息往往藏在：
搜索结果的第 2、3、4 页
视频或笔记的完整互动数据
创作者主页和历史作品
评论区里的价格、预约、体验和真实需求
不同平台对同一话题的反馈差异
FeedSonar 把这些能力包装成 Agent 能理解、能组合、能自动翻页的结构化工具。
示例任务：
“搜索 12 个北京旅游视频，不够就继续翻页；筛选互动高的内容，再读取评论区，总结游客最关心的问题。”
Agent 会自动完成：
关键词搜索 → 判断 has_more → 自动翻页 → 内容详情
→ 作者资料 → 历史作品 → 评论 / 二级评论 → 分析Related MCP server: rednote-analyzer-mcp
能力 | 普通网页搜索 | FeedSonar |
返回网页链接与摘要 | ✓ | ✓ |
返回结构化互动指标 | — | ✓ |
自动翻页满足用户数量 | — | ✓ |
获取作者资料与作品 | — | ✓ |
获取一级 / 二级评论 | — | ✓ |
跨平台统一数据结构 | — | ✓ |
无需注册，无需用户 API Key。选择你的 Agent，复制对应配置即可。
Agent / 客户端 | 推荐连接方式 |
Cursor | 远程 MCP URL / 一键安装 |
Claude Code |
|
Claude Desktop | npm stdio 桥接器 |
Codex CLI / IDE / ChatGPT Desktop |
|
Gemini CLI |
|
VS Code / GitHub Copilot |
|
Windsurf | Cascade 远程 MCP |
Cline | Streamable HTTP |
Roo Code、Cherry Studio 等 | 通用 npm stdio 配置 |
最快方式：一键安装到 Cursor
手动安装：保存到项目的 .cursor/mcp.json，然后重新加载 Cursor。
{
"mcpServers": {
"feedsonar": {
"url": "https://www.aisocialdata.xyz/mcp"
}
}
}claude mcp add --transport http --scope user feedsonar https://www.aisocialdata.xyz/mcp检查连接：
claude mcp listClaude Desktop 使用 stdio 桥接器。要求 Node.js 18 或更高版本。
配置文件：
Windows：%APPDATA%\Claude\claude_desktop_config.json
macOS：~/Library/Application Support/Claude/claude_desktop_config.json
{
"mcpServers": {
"feedsonar": {
"command": "npx",
"args": ["-y", "feedsonar-mcp"]
}
}
}保存后完全退出并重新打开 Claude Desktop。
Codex CLI、IDE 扩展和 ChatGPT Desktop 共用 Codex MCP 配置。添加到 ~/.codex/config.toml；也可在受信任项目中使用 .codex/config.toml。
[mcp_servers.feedsonar]
url = "https://www.aisocialdata.xyz/mcp"重新启动客户端后运行 /mcp 检查工具。
命令行安装：
gemini mcp add --transport http feedsonar https://www.aisocialdata.xyz/mcp或者添加到 ~/.gemini/settings.json：
{
"mcpServers": {
"feedsonar": {
"url": "https://www.aisocialdata.xyz/mcp",
"type": "http"
}
}
}保存为工作区的 .vscode/mcp.json：
{
"servers": {
"feedsonar": {
"type": "http",
"url": "https://www.aisocialdata.xyz/mcp"
}
}
}打开 Copilot Chat 的 Agent 模式；也可以从命令面板运行 MCP: Add Server。
在 Settings → Tools → Add Server 中选择远程 HTTP，填入：
https://www.aisocialdata.xyz/mcp使用 Raw Config 时：
{
"mcpServers": {
"feedsonar": {
"serverUrl": "https://www.aisocialdata.xyz/mcp"
}
}
}保存后在 Cascade MCP 面板点击 Refresh。
在 Cline 的 MCP Servers → Remote Servers 中选择 Streamable HTTP，或使用：
{
"mcpServers": {
"feedsonar": {
"type": "streamableHttp",
"url": "https://www.aisocialdata.xyz/mcp",
"disabled": false,
"autoApprove": []
}
}
}只要客户端支持标准 command + args MCP 配置，就可以使用：
{
"mcpServers": {
"feedsonar": {
"command": "npx",
"args": ["-y", "feedsonar-mcp"]
}
}
}检查 npm 桥接器：
npx -y feedsonar-mcp --version然后向 Agent 提问：
使用
get_hot_list获取抖音前 3 个热点，并告诉我每个热点的热度。
安装成功时，客户端应显示以下 6 个工具：
get_hot_list · social_search · get_content_detail
get_user_profile · get_user_posts · get_comments“搜索 20 个机器人相关视频，自动翻页并按互动量整理选题方向。”
“采集这个视频 100 条评论和热门回复，归纳用户抱怨、购买意向和高频问题。”
“找到人工智能领域创作者，读取公开资料和最近作品，筛选适合合作的账号。”
“对比抖音、小红书、Bilibili 和微博对同一旅游话题的内容角度与用户反馈。”
所有公开能力都经过真实端点黑盒测试；未通过验收的能力不会放进 schema。
平台 | 已验证能力 |
抖音 | 热点、内容/用户搜索、自动翻页、详情、用户资料、作品、一级/二级评论 |
小红书 | 真实分享链接笔记详情、作者、互动量、标签、媒体 |
Bilibili | 真实 BVID 详情、作者、互动指标、评论 |
微博 | 真实帖子详情、作者、互动指标 |
适合热点晨报、选题发现和趋势追踪。
返回 pagination.has_more、next_cursor 和翻页指令。Agent 会根据用户要求的总量继续调用，而不是停在第一页。
返回正文、作者、发布时间、互动指标和媒体信息；支持 standard 与 full 两种详细度。
获取公开资料和作品列表；作品支持自动翻页。
支持一级评论、二级评论和 cursor 翻页。抖音单页最多 30 条。
发布前执行了 10 个隔离的真实业务黑盒场景，每个场景都启动一个全新的 stdio MCP 客户端：
10 / 10 场景通过
搜索 4 页得到 12 个去重视频
评论自动翻 2 页得到 60 条去重评论
创作者作品自动翻 2 页得到 10 条作品
验证抖音评论树、小红书笔记、Bilibili 视频与评论、微博帖子
Node.js 18 / 20 真实安装验证通过
Agent
├─ Remote MCP
└─ npm stdio bridge
↓ HTTPS
Free Hosted MCP
↓
Public social-platform data开源包只负责 MCP 协议桥接，不包含私有采集后端或服务器凭据。
桥接器不会读取本地文件。
查询参数会发送到托管服务以完成取数，请勿提交账号密码、Cookie 或私密数据。
免费服务采用公平使用限速；平台风控或上游变化可能导致个别能力暂时不可用。
npm install
npm test # 离线验证 tools/list 与 tools/call
npm run test:journey自定义兼容端点：
FEEDSONAR_MCP_URL=http://127.0.0.1:9100/mcp npx feedsonar-mcp非本机地址必须使用 HTTPS。
开源桥接器使用 MIT License。托管数据服务及其后端不属于本许可证授权范围。
获取公开内容的一级或二级评论。count 是单页数量，抖音最多 30；当用户要求更多且返回 pagination.has_more=true 时，必须保持内容与层级参数，并用 pagination.next_cursor 自动继续调用，直到满足用户数量或无更多结果。抖音一级评论传 content_id，二级评论还需 comment_id；Bilibili 的 content_id 使用详情返回的数字 aid。
| Name | Required | Description | Default |
|---|---|---|---|
| count | No | 当前页返回条数，最大 30；不是总采集量。 | |
| level | No | root 为一级评论，sub 为二级评论。 | root |
| order | No | Bilibili 评论排序；其他平台忽略。 | like |
| cursor | No | 上一页 pagination.next_cursor；自动翻页时原样传回。 | |
| platform | Yes | 内容所属平台。 | |
| verbosity | No | standard 返回适合 Agent 的详细结构；full 返回上游完整原始数据。 | standard |
| comment_id | No | 读取二级评论时使用的父评论 ID。 | |
| content_id | No | 内容 ID。 | |
| share_text | No | 内容分享文案或链接。 |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
无注解，描述承担了行为透明度的全部责任。它揭露了自动分页行为（必须保持参数并继续调用直到满足数量或无更多结果），并提示count是单页数量而非总数。虽未涉及限流、授权或错误处理，但核心行为已较清楚。
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
描述由三句话组成，第一句直述目的，第二句给出分页规则，第三句给出平台差异。每句信息密度高，无冗余，结构清晰且前置核心信息。
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
考虑到无输出schema和无注解，描述覆盖了主要操作场景（分页、平台参数、数量限制）。但未提及xhs平台的特定传参方式，也没有说明返回结构（除分页字段外），因此略有缺口。
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
尽管schema已有100%字段描述，描述额外补充了参数关系与平台条件：count为单页量并区分抖音上限，抖音二级评论需comment_id，Bilibili的content_id使用详情返回的aid。这些语义在schema中未体现，极大帮助agent正确传参。
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
描述明确说明工具用于获取公开内容的一级或二级评论，动词和资源清晰，且与兄弟工具（如内容详情、搜索）区分明显。同时具体指出抖音和Bilibili的平台差异，功能边界明确。
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
描述提供了平台特定的参数使用指南（抖音需content_id/comment_id，Bilibili的content_id使用详情返回的数字aid），以及分页自动继续的触发条件。但未明确说明在何种情况下应选择其他兄弟工具，缺乏明确的排除性指导。
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
获取单条公开内容的详细数据，包括作者、正文和互动指标。传 content_id 或平台分享文案 share_text。
| Name | Required | Description | Default |
|---|---|---|---|
| platform | Yes | 内容所属平台。 | |
| verbosity | No | standard 返回适合 Agent 的详细结构；full 返回上游完整原始数据。 | standard |
| content_id | No | 视频、笔记或帖子的内容 ID。 | |
| share_text | No | 平台分享文案或分享链接。 |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations provided, the description carries the full burden of behavioral disclosure. It mentions 'public content' and returns author/body/metrics, but does not disclose authentication requirements, rate limits, error behavior, or the relationship between content_id and share_text (e.g., what happens if both are provided).
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is extremely concise: two sentences that front-load the tool's purpose and then give parameter guidance. Every word earns its place, with no redundant or verbose information.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Despite schema coverage, the description is incomplete because it says to pass content_id or share_text, but neither is required in the schema, creating ambiguity about whether at least one must be supplied. Additionally, no return format or error scenarios are described, and the verbosity parameter is only in the schema, not the description. This is a significant gap for a tool with four parameters.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema coverage is 100%, providing descriptions for all parameters. The description adds valuable semantics by stating that either content_id or share_text should be passed, clarifying the alternative relationship not explicit in the schema. This goes beyond the baseline.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the tool's purpose: to retrieve detailed data for a single piece of public content, including author, body, and interaction metrics. It distinguishes from sibling tools like get_hot_list (list) and get_user_posts (user posts) by focusing on single content details.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
No guidance is provided on when to use this tool versus alternatives such as social_search or get_comments. The description only explains what the tool does and how to pass parameters, without explicit when-to-use or when-not-to-use instructions.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
获取社媒平台当前热点榜或热搜榜。适合回答今日热点、热门话题和实时趋势；不适合按关键词搜索历史内容。
| Name | Required | Description | Default |
|---|---|---|---|
| limit | No | 返回条数，默认 20。 | |
| platform | Yes | 平台，当前支持 douyin。 |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations provided, the description carries the full burden. It adds context that the data is current/real-time and not historical, which is valuable. However, it does not disclose return format, pagination behavior, rate limits, or any other operational details, leaving some uncertainty for an agent.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is composed of two compact sentences, delivering the core purpose and usage guidance without redundant phrasing or irrelevant detail.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a simple two-parameter tool with no output schema, the description sufficiently explains the purpose and use-case boundaries. It could have added a note on the response structure, but the essential context for invocation is present. The platform restriction is implied by the schema's enum.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The schema already provides 100% coverage of both parameters (platform enum and limit default/range), so the description offers no additional parameter-level semantics. The description's mention of 'current' applies to the tool's behavior, not to parameter specifics. Baseline 3 is appropriate.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the tool retrieves the current hot/trending lists from social media platforms, using the verb '获取' and specifying the resource. It distinguishes itself from sibling search tools by emphasizing 'current' trends, making its unique purpose evident.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
It explicitly states suitable use cases (answering today's hot topics, popular topics, real-time trends) and explicitly excludes keyword-based historical content search, which clearly differentiates it from the sibling social_search tool. However, it does not name the alternative tool directly, so it falls slightly short of full 5.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
获取公开用户发布内容列表。count 是单页目标；当用户要求更多且返回 pagination.has_more=true 时，必须保持用户参数并用 pagination.next_cursor 自动继续调用，直到满足用户数量或无更多结果。
| Name | Required | Description | Default |
|---|---|---|---|
| count | No | 期望返回条数。 | |
| cursor | No | 上一页 pagination.next_cursor；自动翻页时原样传回。 | |
| user_id | No | 平台用户 ID。 | |
| platform | Yes | 用户所属平台。 | |
| verbosity | No | standard 返回适合 Agent 的详细结构；full 返回上游完整原始数据。 | standard |
| share_text | No | 用户主页分享文案或链接。 | |
| sec_user_id | No | 抖音 sec_user_id。 |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations provided, the description carries the full burden. It discloses that count is a per-page target, and that the tool will automatically paginate while preserving user parameters until the requested count is met or results are exhausted. This is meaningful behavioral context beyond the schema.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is two concise sentences, front-loaded with the primary purpose. Every sentence adds value, with no redundant information.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given 7 parameters and no output schema, the description covers the crucial pagination behavior but does not clarify how to identify the user (user_id, share_text, or sec_user_id) or require one of them. However, the schema descriptions for those parameters partially compensate, making the description reasonably complete for the core use case.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is 100%, so the baseline is 3. The description adds semantics for count (single-page target) and clarifies cursor usage in auto-pagination, which goes beyond the schema's field descriptions. Other parameters are adequately covered by the schema.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the tool retrieves a list of a user's public posts, using a specific verb and resource. It distinguishes itself from siblings like get_user_profile and get_comments by focusing on content list retrieval.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description provides a clear pagination protocol (auto-continue with next_cursor when has_more=true), but it does not explicitly mention when to use this tool over alternatives like social_search or get_content_detail. There are no exclusions or alternative tool references.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
获取公开用户资料。传 user_id、抖音 sec_user_id 或用户分享文案；不要用于获取私密账号数据。
| Name | Required | Description | Default |
|---|---|---|---|
| user_id | No | 平台用户 ID。 | |
| platform | Yes | 用户所属平台。 | |
| verbosity | No | standard 返回适合 Agent 的详细结构；full 返回上游完整原始数据。 | standard |
| share_text | No | 用户主页分享文案或链接。 | |
| sec_user_id | No | 抖音 sec_user_id。 |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations provided, the description carries the full burden of behavioral disclosure. It states that the tool retrieves public profiles and implicitly warns against private data, but it does not disclose return format, pagination, rate limits, error behavior, or how the tool handles missing/invalid identifiers. The input guidance is present, but behavioral transparency beyond the basic function is minimal.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is two short sentences in Chinese. It front-loads the core purpose, then gives input options and a usage warning. There is no filler or repetition of schema details. Every clause earns its place.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
There is no output schema, and the description does not explain what fields or structure the returned profile contains, nor any platform-specific variations. It also does not describe error conditions or how the verbosity parameter affects output. Despite the schema covering parameters, the description leaves the agent under-informed about the tool's actual return value and edge cases.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is 100% (all parameters have descriptions), so baseline is 3. The description adds value by clarifying that user_id, sec_user_id, and share_text are alternative ways to identify the user, and that sec_user_id is Douyin-specific. This supplements the schema's individual parameter descriptions and helps the agent choose among them.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the tool's purpose as '获取公开用户资料' (get public user profile), which is a specific verb+resource combination. It also notes that it accepts multiple identifier types (user_id, sec_user_id, share text), further clarifying scope. However, it does not explicitly reference sibling tools like get_user_posts or social_search, so it lacks explicit sibling differentiation.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description provides a clear context: this tool is for public user profiles, with a direct instruction on what to pass (user_id, sec_user_id, or share text). It also includes a when-not directive ('不要用于获取私密账号数据' – do not use for private account data). However, it does not mention alternative tools for other use cases (e.g., getting user posts or comments), so it stops short of full alternatives guidance.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Each tool targets a distinct resource and action—hot lists, keyword search, content details, user profiles, user posts, and comments—with no conceptual overlap. Even social_search and get_user_posts differ by keyword-based discovery vs. user-specific listing.
Five of six tools follow the get_<object> pattern, creating a predictable structure. The single exception, social_search, breaks the pattern but remains clear and readable, so a mild deviation.
With 6 tools, the server is well-scoped for social media data retrieval, covering discovery, detail, and user-centric queries without excess or deficiency.
The toolset forms a complete read-only workflow for social media data: hot list for trends, search for discovery, content detail for specific items, user profile and posts for user-centric exploration, and comments for engagement context. No obvious missing operation within this domain.
Unclaimed servers have limited discoverability.
If you are the server author, to access and configure the admin panel.
MCP server: AI-agent access to Chinese social & trend signals — Douyin, Weibo, Xiaohongshu/RedNote,
Remote MCP server for China brand visibility, destination demand, and KOL discovery workflows.
MCP server for QPost — lets AI agents publish video and image posts to YouTube, TikTok, Instagram.
MCP server connecting AI agents to 100+ apps (Gmail, Slack, Notion, GitHub) via one-click OAuth.
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/yy-MAX-HUB/feedsonar-mcp'
If you have feedback or need assistance with the MCP directory API, please join our Discord server
social_searchA
按关键词搜索抖音公开内容或用户。count 是单页目标；当用户要求的总量超过本页且返回 pagination.has_more=true 时，必须传回 pagination.next_cursor、search_id、search_context 自动继续调用，直到满足用户数量或无更多结果，不能只返回第一页。
TDQS
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations, the description carries the full burden and excels by disclosing the critical auto-pagination behavior and prohibiting returning only the first page. This is essential for correct agent behavior and goes well beyond what the schema provides.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is a single dense, front-loaded sentence that states the purpose, defines the key parameter semantics, and specifies the pagination rule without any redundancy or fluff. It earns every word.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
The tool is complex with 8 parameters and no output schema. The description thoroughly covers the non-obvious pagination algorithm, which is the main risk for misuse. It does not describe the result item structure, but that is less critical and inferable from the purpose and sibling detail tools.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Although the schema names all parameters with descriptions (100% coverage), the description adds crucial meaning: it clarifies that 'count' is the per-page target, not the user's total, and mandates echoing three pagination tokens together. This adds value beyond the schema fields.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states '按关键词搜索抖音公开内容或用户' (search Douyin public content or users by keyword), identifying a specific verb and resource. This distinguishes it from sibling getter tools like get_content_detail and get_user_posts.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description gives explicit pagination instructions: count is per-page, and when has_more=true, the agent must pass back next_cursor, search_id, and search_context to continue until the user's total is met or no more results. It does not explicitly mention alternatives, but the search-vs-get distinction is clear from tool names and purpose.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.