Markdown Fetch

Optimizes web fetching by using Cloudflare's Markdown for Agents, reducing token consumption by ~80%

Security Scan

The skill does what it says: it adds an Accept header to prefer Markdown responses and provides simple fetch helpers — there are no unexpected credentials, installs, or hidden endpoints.

Purpose & Capability

The name/description (request Markdown to save tokens) matches the code and instructions. The files only implement an optimizedFetch, batchFetch, and simple HTML fallback; nothing requests unrelated capabilities or credentials.

Instruction Scope

SKILL.md recommends finding and updating HTTP requests (fetch/axios/request) to add the Accept header and to treat text/markdown responses specially. That is within the stated purpose, but it implies a code-wide change — reviewers should manually confirm changes rather than blindly applying them across all requests because altering every request can change behavior for endpoints that expect other Accept headers.

Install Mechanism

There is no install spec and no external downloads; this is instruction + small JS module only, so nothing will be written to disk by an installer. Low risk from installation mechanisms.

Credentials

The skill requests no environment variables, no credentials, and no config paths. The code makes outgoing HTTP requests to whatever URL you call it with; there are no hidden or hard-coded endpoints or secret-access calls.

Persistence & Privilege

The skill is not marked always:true and does not modify other skills or system settings. It contains only library functions and a test harness — no persistent or privileged behavior is requested.

Assessment

This skill is internally consistent and small: it simply sets an Accept header and returns markdown when available. Before adopting it broadly: 1) Review and apply its changes manually (don't automatically rewrite every HTTP call) because changing Accept headers can alter server behavior; 2) Test in staging — not all sites support text/markdown and behavior may differ; 3) Verify your runtime provides a global fetch (or adapt to node-fetch/axios) so nothing breaks; 4) Note the skill will perform network fetches to URLs you pass it — ensure you don't unintentionally fetch internal endpoints. No secrets are requested by the skill, and there are no hidden endpoints in the code.

Markdown Fetch - 网页抓取优化

背景

Cloudflare 推出 Markdown for Agents 功能：

• AI 请求时返回 Markdown 格式
• Token 消耗比 HTML 减少约 80%

使用方法

在需要网页抓取时，使用优化后的 fetch 函数：

const { optimizedFetch } = require('./markdown-fetch');

const result = await optimizedFetch('https://example.com');
// result.markdown - Markdown 内容（如果有）
// result.html - HTML 内容（备用）
// result.tokensSaved - 节省的 tokens（如果有）

核心逻辑

async function optimizedFetch(url, options = {}) {
  const headers = {
    'Accept': 'text/markdown, text/html',
    ...options.headers
  };

  const response = await fetch(url, { ...options, headers });
  
  const contentType = response.headers.get('content-type');
  const xMarkdownTokens = response.headers.get('x-markdown-tokens');
  
  let result = {
    url,
    contentType,
    tokensSaved: xMarkdownTokens ? parseInt(xMarkdownTokens) : null
  };
  
  if (contentType.includes('text/markdown')) {
    result.markdown = await response.text();
    result.format = 'markdown';
  } else {
    result.html = await response.text();
    result.format = 'html';
  }
  
  return result;
}

响应处理

Content-Type	处理方式
text/markdown	直接使用，跳过 HTML 解析
text/html	走原有解析逻辑

可选：x-markdown-tokens 日志

如果响应中有 x-markdown-tokens header，记录到日志：

if (result.tokensSaved) {
  console.log(`[Markdown Fetch] Token 节省: ${result.tokensSaved}`);
}

改动范围

1. 找到所有 HTTP 请求（fetch/axios/request）
2. 统一添加 header
3. 响应处理加判断

测试验证

找一个 Cloudflare 托管的网站测试：

curl -H "Accept: text/markdown, text/html" https://cloudflare-example.com

确认收到 content-type: text/markdown 响应。