← 返回 Skills 市场

cf markdown agents

Name: cf markdown agents
Author: xinian5216

作者 xinian5216 · GitHub ↗ · v1.0.0 · MIT-0

cross-platform ✓ 安全检测通过

227

总下载

当前安装

版本数

在 OpenClaw 中安装

/install cf-markdown-agents

功能描述

Fetch web content using Cloudflare's Markdown for Agents protocol. Automatically converts HTML to Markdown with 80% token reduction. Use when fetching web pa...

使用说明 (SKILL.md)

Cloudflare Markdown for Agents

This skill enables fetching web content using Cloudflare's Markdown for Agents protocol, which automatically converts HTML pages to clean, structured Markdown format.

What is Markdown for Agents?

Cloudflare's Markdown for Agents is a content negotiation feature that:

Automatically converts HTML to Markdown at the edge
Reduces token usage by ~80% compared to raw HTML
Returns clean, structured content perfect for AI processing
Adds x-markdown-tokens header showing estimated token count
Includes Content-Signal headers for AI usage permissions

Usage

Basic Fetch

Use the provided script to fetch any URL with Markdown for Agents support:

scripts/fetch-markdown.sh \x3CURL>

Example:

scripts/fetch-markdown.sh "https://developers.cloudflare.com/agents/"

In Code

TypeScript/JavaScript example:

const response = await fetch("https://example.com/page", {
  headers: {
    Accept: "text/markdown, text/html",
  },
});

const tokenCount = response.headers.get("x-markdown-tokens");
const markdown = await response.text();

cURL example:

curl https://example.com/page -H "Accept: text/markdown"

Response Headers

content-type: text/markdown; charset=utf-8 - Content is Markdown
x-markdown-tokens: \x3Cnumber> - Estimated token count
content-signal: ai-train=yes, search=yes, ai-input=yes - Usage permissions

Supported Sites

Any site using Cloudflare with Markdown for Agents enabled:

Cloudflare Developer Documentation (developers.cloudflare.com)
Cloudflare Blog (blog.cloudflare.com)
Any site with the feature enabled in Cloudflare dashboard

Benefits

Aspect	HTML	Markdown
Tokens	16,180	3,150
Reduction	-	~80%
Structure	Complex	Clean
AI Parsing	Hard	Easy

References

Full protocol details: references/protocol.md
API examples: references/examples.md

安全使用建议

This skill appears coherent and does what it says: send requests with Accept: text/markdown and return the Markdown response and headers. It does not request credentials. Before enabling, consider: (1) the skill will fetch any URL you provide — if the agent is allowed to run autonomously it could be used to reach internal services (SSRF/internal network exposure) or fetch sensitive URLs, so restrict allowed targets or monitor requests; (2) example code shows how to add API tokens (optional) — you should never add secrets unless you trust the code; (3) the provided script prints headers and body (including any returned sensitive content) so treat outputs as potentially sensitive. If you only need occasional manual fetches, run the script locally rather than granting broad autonomous agent access.

功能分析

Type: OpenClaw Skill Name: cf-markdown-agents Version: 1.0.0 The skill is a legitimate utility designed to fetch web content using Cloudflare's 'Markdown for Agents' protocol to reduce token usage for AI agents. It consists of a simple shell script (scripts/fetch-markdown.sh) that wraps curl with appropriate headers and comprehensive documentation (SKILL.md, references/examples.md) explaining the protocol and providing implementation examples. No evidence of data exfiltration, malicious execution, or prompt injection was found.

能力评估

✓ Purpose & Capability

Name/description (fetch pages via Cloudflare Markdown for Agents) align with the included script, examples, and documentation. The included curl/fetch/worker examples are exactly what you'd expect for this purpose.

ℹ Instruction Scope

Runtime instructions and the shell script only perform HTTP requests with an Accept header for text/markdown and print response headers/body. Examples include optional usages that mention Authorization tokens or Cloudflare Workers/AI helpers — these are illustrative and not required by the skill, but they show how a user could add credentials if they choose.

✓ Install Mechanism

No install spec; this is instruction-only with a small shell script. Nothing is downloaded from remote URLs or written to disk beyond the provided files.

✓ Credentials

The skill declares and requires no environment variables, credentials, or config paths. Example snippets show how one might use API tokens if they wrote additional code, but the skill itself does not request secrets.

✓ Persistence & Privilege

always is false and the skill does not attempt to modify other skills or system-wide agent settings. It does not request persistent privileges.

如何使用

确保已安装 OpenClaw（本地或 Docker 部署）
在对话框中输入安装命令：/install cf-markdown-agents
安装完成后，直接呼叫该 Skill 的名称或使用 /cf-markdown-agents 触发
根据 Skill 的参数说明提供必要输入，即可获得结构化输出

版本历史

v1.0.0

Initial release of cf-markdown-agents. - Enables fetching web content using Cloudflare's Markdown for Agents protocol. - Automatically converts HTML pages to clean, structured Markdown format with ~80% token reduction. - Provides examples for usage with shell scripts, TypeScript/JavaScript, and cURL. - Describes response headers including token count and AI usage permissions. - Lists supported sites and summarizes key benefits for AI processing.

元数据

Slug cf-markdown-agents

版本 1.0.0

许可证 MIT-0

累计安装 0

当前安装数 0

历史版本数 1

常见问题