UAParse
Parse a User-Agent: browser, OS, device, plus bot, search crawler, AI crawler and AI agent flags.
我该使用它吗
质量与安全性
基于对工具定义和协议合规性的自动分析。
上下文开销
这是每次将服务器的工具加载到模型上下文窗口时所消耗的大致 token 数。数值越高，可用于其他任务的注意力就越少。
安装
一键安装
将以下内容添加到你的 `claude_desktop_config.json` 文件中：
{
"mcpServers": {
"uaparse": {
"url": "https://uaparse.mike-tusa.workers.dev/mcp"
}
}
}远程端点
https://uaparse.mike-tusa.workers.dev/mcpstreamable-http它能做什么
工具清单
工具（1）
🟢parse_user_agent(ua, include_raw)
Parse one User-Agent string. Returns browser, engine, OS, device type and CPU (ua-parser-js 1.0.41), and whether it is a known bot, search crawler, AI crawler (with its AI purpose) or an AI agent acting for a user (is_ai_agent, ai_agent_kind fetch | browsing), plus the bot name, operator, robots.txt token and operator docs link from a curated list of 151 bots (list 2026-10-08). verified is always false: the User-Agent is self-reported and UAParse does no IP or reverse-DNS checks. Read-only, no network lookups. Same engine and limits as the UAParse JSON API (POST /v1/parse); each call is one lookup. UAParse is independent and not affiliated with Meta, OpenAI, Google, Anthropic, Amazon, Perplexity or Mistral AI.
输入模式
{
"type": "object",
"properties": {
"ua": {
"type": "string",
"minLength": 1,
"maxLength": 2048,
"description": "The User-Agent string to parse, exactly as sent in the User-Agent header."
},
"include_raw": {
"type": "boolean",
"default": false,
"description": "Also return the full POST /v1/parse JSON in structuredContent.raw. Larger output."
}
},
"required": [
"ua"
],
"additionalProperties": false
}输出模式
{
"type": "object",
"properties": {
"ua": {
"type": "string",
"description": "The parsed User-Agent (clipped to 300 characters; see ua_length)"
},
"ua_length": {
"type": "integer"
},
"browser": {
"type": "object",
"properties": {
"name": {
"type": "string"
},
"version": {
"type": "string"
},
"major": {
"type": "string"
}
}
},
"engine": {
"type": "object",
"properties": {
"name": {
"type": "string"
},
"version": {
"type": "string"
}
}
},
"os": {
"type": "object",
"properties": {
"name": {
"type": "string"
},
"version": {
"type": "string"
}
}
},
"device": {
"type": "object",
"properties": {
"type": {
"type": "string",
"description": "desktop, mobile, tablet, smarttv, console, wearable, xr, embedded, bot or unknown"
},
"vendor": {
"type": "string"
},
"model": {
"type": "string"
}
},
"required": [
"type"
]
},
"cpu": {
"type": "object",
"properties": {
"architecture": {
"type": "string"
}
}
},
"is_bot": {
"type": "boolean"
},
"is_ai_crawler": {
"type": "boolean"
},
"is_ai_agent": {
"type": "boolean",
"description": "true only for User-Agents the operator documents as acting live for an individual user"
},
"is_search_crawler": {
"type": "boolean"
},
"bot": {
"type": "object",
"description": "Present when is_bot is true. Fields that are unknown are omitted.",
"properties": {
"name": {
"type": "string"
},
"operator": {
"type": "string"
},
"category": {
"type": "string"
},
"category_label": {
"type": "string"
},
"ai_purpose": {
"type": "string",
"enum": [
"training",
"search",
"user_fetch",
"agent"
]
},
"ai_agent_kind": {
"type": "string",
"enum": [
"fetch",
"browsing"
]
},
"robots_token": {
"type": "string"
},
"token_only": {
"type": "boolean"
},
"info_url": {
"type": "string",
"description": "Operator documentation"
},
"match": {
"type": "string",
"enum": [
"list",
"heuristic"
]
},
"verified": {
"type": "boolean",
"description": "Always false"
}
},
"required": [
"category",
"match",
"verified"
]
},
"summary": {
"type": "string",
"description": "One-line plain-English verdict"
},
"list_version": {
"type": "string"
},
"parser": {
"type": "string"
},
"cached": {
"type": "boolean"
},
"tier": {
"type": "string",
"enum": [
"anonymous",
"free",
"pro"
]
},
"version": {
"type": "string"
},
"raw": {
"type": "object",
"description": "Full POST /v1/parse response (only when include_raw is true)"
}
},
"required": [
"ua",
"ua_length",
"browser",
"engine",
"os",
"device",
"cpu",
"is_bot",
"is_ai_crawler",
"is_ai_agent",
"is_search_crawler",
"summary",
"list_version",
"version"
]
}社区
证据