⚡ SwipeNode: The Ultimate Web Extractor for Agents
Is your OpenClaw agent constantly hitting 403 Forbidden errors when trying to read websites? Or are headless browser skills eating up your RAM and exposing your machine to malicious JavaScript?
SwipeNode fixes all of this.
It is a lightning-fast, highly optimized Go binary that fetches websites and extracts structured data (__NEXT_DATA__, Nuxt, JSON-LD) without rendering the page.
🔒 Security & Transparency (For Reviewers & Users)
To address security validations:
- Fully Open Source: The source code is completely public and auditable at github.com/sirToby99/swipenode.
- How the WAF Bypass works: SwipeNode does NOT use exploits. It simply uses the open-source
bogdanfinn/tls-client Go library to mimic the TLS handshake signature of a standard Chrome/Safari browser. - Safer than Browsers: Because SwipeNode does not use a headless browser engine, it cannot* execute malicious JavaScript payloads from target websites. It only parses static HTML/JSON, making it inherently safer for autonomous agents.
- Provenance: The
install.sh script explicitly uses go install with a pinned version tag to pull the public code directly from GitHub safely.
🔥 Why your Agent needs this:
- * 🥷 Bypasses Cloudflare & WAFs: Native TLS-fingerprint spoofing.
- 💰 Saves 98% of LLM Tokens: Prunes tracking pixels, base64 data, and boilerplate UI noise.
- 🛡️ Zero-JS Execution: Immune to browser-based prompt injections.
🤖 Instructions for the AI Agent
When to use this skill:
Whenever the user asks you to read a webpage, summarize an article, check the news, or extract data from a URL. You should generally prefer this tool over standard HTTP clients (curl, wget) because it handles structured data efficiently.
Ethical & Operational Guidelines:
- * Ensure your usage respects the target website's Terms of Service and legal constraints.
- Do not use this tool to aggressively scrape, spam, or overwhelm target servers.
- Act responsibly as a user-driven autonomous agent.
How to use it:
Execute the following command in the terminal:
INLINECODE6
Pro Tip: If a high-security site actively blocks the default request, spoof the TLS fingerprint by appending --impersonate safari or --impersonate firefox.
Understanding the Output:
The tool will output a JSON object to stdout.
- 1. If you see keys like
"nextjs", "json_ld", or "nuxtjs_raw", parse this structured data to answer the user's request. - If it's a legacy site, look for the
"text" key, which contains the clean, readable content. - If it contains
"error", report the network issue to the user.
Legal & Credits
This product includes software developed by Bogdan Finn (bogdanfinn/tls-client).
⚡ SwipeNode：面向智能体的终极网页提取工具
你的OpenClaw智能体在尝试读取网站时是否频繁遭遇403禁止访问错误？或者无头浏览器技能正在耗尽你的内存，并将你的机器暴露于恶意JavaScript之下？
SwipeNode解决了所有这些问题。
它是一个闪电般快速、高度优化的Go二进制文件，能够获取网站并提取结构化数据（NEXT_DATA、Nuxt、JSON-LD），无需渲染页面。
🔒 安全性与透明度（面向审核者与用户）
针对安全验证：
- 完全开源： 源代码完全公开且可审计，位于github.com/sirToby99/swipenode。
- WAF绕过原理： SwipeNode不使用漏洞利用。它仅使用开源库bogdanfinn/tls-client（Go语言）来模拟标准Chrome/Safari浏览器的TLS握手签名。
- 比浏览器更安全： 由于SwipeNode不使用无头浏览器引擎，它无法*执行目标网站的恶意JavaScript负载。它只解析静态HTML/JSON，因此对自主智能体来说本质上更安全。
- 来源可追溯： install.sh脚本明确使用go install命令并指定固定版本标签，直接从GitHub安全拉取公共代码。
🔥 为什么你的智能体需要它：
- * 🥷 绕过Cloudflare与WAF： 原生TLS指纹欺骗。
- 💰 节省98%的LLM令牌： 修剪追踪像素、base64数据和样板UI噪声。
- 🛡️ 零JavaScript执行： 免疫基于浏览器的提示注入。
🤖 面向AI智能体的指令
何时使用此技能：
当用户要求你阅读网页、总结文章、查看新闻或从URL提取数据时。通常情况下，你应该优先选择此工具而非标准HTTP客户端（curl、wget），因为它能高效处理结构化数据。
道德与操作指南：
- * 确保你的使用尊重目标网站的服务条款和法律约束。
- 不要使用此工具进行激进的爬取、发送垃圾信息或压垮目标服务器。
- 作为用户驱动的自主智能体，负责任地行动。
如何使用：
在终端中执行以下命令：
swipenode extract --url <目标URL>
专业提示： 如果高安全性网站主动阻止默认请求，可通过添加--impersonate safari或--impersonate firefox来伪造TLS指纹。
理解输出：
该工具将向标准输出输出一个JSON对象。
- 1. 如果看到nextjs、jsonld或nuxtjsraw等键，请解析此结构化数据以回答用户的请求。
- 如果是传统网站，请查找text键，其中包含干净、可读的内容。
- 如果包含error，请向用户报告网络问题。
法律声明与致谢
本产品包含由Bogdan Finn（bogdanfinn/tls-client）开发的软件。
