Tooltrace

io.tooltrace由 tooltrace.io 發布官方發布者v0.1.3更新於 Nov 2, 2026

Web tools for AI agents: scrape pages to Markdown, audit SEO, detect tech stacks, check sitemaps

已驗證Streamable HTTP可網頁執行網頁搜尋與抓取資料與分析

概覽

AI 產生的概覽

ToolTrace 擷取網頁，讓助理取得乾淨的 Markdown、中繼資料、連結、JSON-LD、SEO 稽核、技術堆疊偵測與網站地圖資料。

可以做什麼
  • 將網頁主內容擷取為 Markdown、純文字、中繼資料、連結、JSON-LD 與內容分段
  • 讀取頁面中繼資料，例如標題、描述、標準網址、Open Graph 與 Twitter 卡片欄位
  • 列出所有內部與外部連結，包含錨點文字與正規化網址
  • 擷取 JSON-LD 結構化資料，例如 Article、Product、Organization、FAQ 等實體
  • 執行附評分的頁面 SEO 稽核，提供檢查項目、佐證與建議
  • 偵測網站的 CMS、框架、分析工具、CDN、主機等技術，並檢視 XML 網站地圖
適用情境
適合助理需要讀取或分析公開網頁的情況：擷取內容供 RAG 匯入、查看競品技術堆疊、稽核到達頁 SEO，或透過網站地圖盤點網站結構。
執行需求
需要 ToolTrace API 金鑰：本機 npm 套件透過 TOOLTRACE_API_KEY 環境變數提供，託管端點則透過 Authorization bearer 標頭提供。需要遠端端點，或以 Node.js 透過 stdio 執行 npm 套件。需能連線 ToolTrace API。
安裝前請注意
API 金鑰屬機密且會傳送給 ToolTrace；本機 stdio 模式金鑰留在自己電腦，託管模式則經 Authorization 標頭傳輸。呼叫會消耗計量方案的額度，瀏覽器渲染比靜態抓取更貴。送出的網址由 ToolTrace 伺服器擷取，請勿送出私有或內部網址。

安裝

在 SourceWeft 中

  1. 開啟 儀表板中的 Tooltrace，將其新增到工作區。
  2. 連接 Authorization token 憑證。它們會被私密儲存，絕不會在公開頁面上展示。
  3. 為需要使用其工具的對話啟用該服務。

Web executable，透過 Streamable HTTP。 遠端服務在工作區中設定後即可從網頁執行環境執行。

交給代理程式安裝

把這段提示詞貼上到 Claude Code、Codex、Cursor 等代理程式，或 SourceWeft 對話中。代理程式會讀取該伺服器的安裝指南，先向你展示發布者、端點和登入方式，經你同意後再把它新增到你的客戶端。金鑰不會寫進任何命令。

閱讀 https://sourceweft.com/mcp/io.tooltrace%2Fmcp-server/install.md，並按其中說明把這個 MCP 伺服器接入我的客戶端。修改任何設定之前，先告訴我它的發布者、端點和登入方式，等我確認。不要把 API 金鑰寫進命令或我會提交的檔案；問我金鑰應從哪裡取得。

檢視代理程式所依據的安裝指南

其他 MCP 客戶端

把它新增到你客戶端的 mcpServers 設定中。

{
  "mcpServers": {
    "tooltrace": {
      "type": "http",
      "url": "https://mcp.tooltrace.io/mcp",
      "headers": {
        "Authorization": "Bearer <YOUR_TOKEN>"
      }
    }
  }
}

README

ToolTrace MCP Server

Give AI agents a complete web intelligence toolkit. This Model Context Protocol server connects Claude Code, Claude Desktop, Cursor, VS Code, Codex, Windsurf and any other MCP client to the ToolTrace API.

One tool per ToolTrace endpoint: scrape any page to clean Markdown, read metadata and JSON-LD, extract links, run on-page SEO audits, detect a site's tech stack, and check XML sitemaps. Free tier included, no card required.

Setup guide: tooltrace.io/mcp

Tools

ToolDescription
tooltrace_extractExtract clean Markdown, text, metadata, links, JSON-LD, and sections from any webpage
tooltrace_metadataGet page title, description, canonical URL, Open Graph, and Twitter card fields
tooltrace_linksExtract all internal and external links with anchor text
tooltrace_schemaExtract JSON-LD structured data (Article, Product, FAQ, etc.)
tooltrace_seo_auditRun an SEO audit with scored checks and recommendations
tooltrace_tech_stackDetect CMS, frameworks, analytics, CDN, hosting, and more
tooltrace_sitemapInspect and parse XML sitemaps

Two ways to connect

Hosted, for clients that take a connector URL. Nothing to install or update:

https://mcp.tooltrace.io/mcp
bash
claude mcp add --transport http tooltrace https://mcp.tooltrace.io/mcp \  --header "Authorization: Bearer your-api-key"

Also listed on Smithery.

Local, over stdio, for clients that run a process. Use this when your client does not speak HTTP, or when you would rather your API key never left your own machine. That is the quick start below.

Both expose the same tools and bill the same credits.

Quick start

1. Get a free API key

Sign up at tooltrace.io/signup. The free plan includes 1,000 credits per month.

2. Install

bash
npm install -g @tooltrace/mcp-server

3. Configure your MCP client

Claude Desktop (~/Library/Application Support/Claude/claude_desktop_config.json):

json
{  "mcpServers": {    "tooltrace": {      "command": "tooltrace-mcp",      "env": {        "TOOLTRACE_API_KEY": "your-api-key"      }    }  }}

Claude Code:

bash
claude mcp add tooltrace tooltrace-mcp -e TOOLTRACE_API_KEY=your-api-key

Cursor (.cursor/mcp.json):

json
{  "mcpServers": {    "tooltrace": {      "command": "tooltrace-mcp",      "env": {        "TOOLTRACE_API_KEY": "your-api-key"      }    }  }}

VS Code (.vscode/mcp.json):

json
{  "servers": {    "tooltrace": {      "command": "tooltrace-mcp",      "env": {        "TOOLTRACE_API_KEY": "your-api-key"      }    }  }}

Codex (~/.codex/config.toml):

toml
[mcp_servers.tooltrace]command = "tooltrace-mcp"env = { TOOLTRACE_API_KEY = "your-api-key" }

Windsurf (mcp_config.json):

json
{  "mcpServers": {    "tooltrace": {      "command": "tooltrace-mcp",      "env": {        "TOOLTRACE_API_KEY": "your-api-key"      }    }  }}

Any other MCP client: run tooltrace-mcp over stdio and set the TOOLTRACE_API_KEY environment variable. Full per-client setup guide at tooltrace.io/mcp.

Run with npx (no install)

bash
TOOLTRACE_API_KEY=your-api-key npx @tooltrace/mcp-server

Usage examples

Once connected, your AI agent can:

  • "Extract the main content from this blog post as Markdown"
  • "What technologies does competitor.com use?"
  • "Run an SEO audit on our landing page"
  • "Get all the links from this documentation page"
  • "Extract the JSON-LD schema from this product page"
  • "Check the sitemap for this website"

Rendering modes

All tools support a render parameter:

  • never: Fast static fetch, 1 credit. Best for server-rendered pages.
  • auto (default): Starts static, renders in browser if needed. 1 or 5 credits.
  • always: Full browser rendering, 5 credits. For JavaScript-heavy SPAs.

Development

bash
git clone https://github.com/ToolTrace-io/mcp-server.gitcd mcp-servernpm installnpm run buildTOOLTRACE_API_KEY=your-key node dist/index.js

License

MIT

來源：README.md，提交 0692e7e

工具

7

tooltrace_extract

Extract webpage content

Extract clean content from a webpage. Returns Markdown, plain text, metadata, links, JSON-LD schema, and content sections. Use this for scraping, RAG ingestion, or content analysis.

輸入參數url必填moderenderincludewait_untilwait_for_selector

tooltrace_metadata

Read page metadata

Extract page metadata: title, description, canonical URL, author, publication date, favicon, Open Graph, and Twitter card fields. Lightweight alternative to full extraction.

輸入參數url必填renderwait_untilwait_for_selector

tooltrace_schema

Extract JSON-LD structured data

Extract JSON-LD structured data from a webpage. Returns schema.org entities like Article, Product, Organization, FAQ, BreadcrumbList, etc.

輸入參數url必填moderenderwait_untilwait_for_selector

tooltrace_seo_audit

Audit on-page SEO

Run an SEO audit on a webpage. Returns a score (0-100), weighted checks for metadata, headings, images, canonical signals, robots directives, social tags, schema, and content length, with evidence and recommendations.

輸入參數url必填renderwait_untilwait_for_selector

tooltrace_tech_stack

Detect the tech stack behind a site

Detect the technology stack of a website: CMS, frameworks, JavaScript libraries, analytics, CDN, hosting, fonts, security tools, and more. Returns categorized detections with confidence levels.

輸入參數url必填renderwait_untilwait_for_selector

tooltrace_sitemap

Check an XML sitemap

Inspect a website's sitemap. Discovers sitemap URLs, parses sitemap XML, and returns listed page URLs with last-modified dates and change frequencies.

輸入參數url必填render

版本歷史

1
  1. v0.1.3最新Sep 16, 2026

評分

已滿足適用檢查項的 100%

該評分衡量此收錄條目的完整程度，依據註冊中心條目、端點與公開來源計算得出。它不是對伺服器本身的評價，也不表示伺服器安全或目前可用：可連通性是帶日期的檢查結果，評分也不包含安裝量或使用者評價。

  • 近 30 天內可連通必需25 分
  • 已宣告認證方式必需10 分
  • 已列出工具必需15 分
  • 描述必需15 分
  • 圖示10 分
  • 文件15 分
  • 授權條款10 分