← Back to Skills Marketplace
Jina Web Fetcher - 网页抓取
by
xaiohuangningde
· GitHub ↗
· v1.0.0
1669
Downloads
0
Stars
10
Active Installs
1
Versions
Install in OpenClaw
/install jina-web-fetcher
Description
使用 Jina AI 抓取网页内容，绕过搜索引擎限制。支持任意URL，支持 Google 搜索结果抓取。
README (SKILL.md)
Jina Web Fetcher
使用 Jina AI 服务抓取网页内容。
安装
无需安装，直接使用 curl。
使用
# 抓取任意网页
curl -s "https://r.jina.ai/http://目标URL"
# 抓取 Google 搜索结果
curl -s "https://r.jina.ai/http://www.google.com/search?q=搜索词"
示例
# 抓取 GitHub Trending
curl -s "https://r.jina.ai/http://github.com/trending"
# 抓取 Hacker News
curl -s "https://r.jina.ai/http://news.ycombinator.com"
# 抓取特定文章
curl -s "https://r.jina.ai/http://example.com/article"
注意
- Google 被封锁时使用 r.jina.ai 仍可能被拦截
- 大部分网站都可以正常抓取
Usage Guidance
This skill simply tells the agent to use r.jina.ai as a proxy to fetch any URL. Before installing or using it consider: 1) Privacy/exfiltration — any URL and page contents you ask to fetch will be relayed to r.jina.ai (including potentially sensitive internal endpoints). 2) SSRF risk — you can inadvertently expose internal services by fetching internal URLs through the proxy. 3) Legal/ToS risk — using a proxy to 'bypass search engine restrictions' may violate site terms of service. 4) If you must use it, avoid sending credentials or private/internal URLs, and verify r.jina.ai's privacy policy or run your own trusted proxy instead. If you need simple public-page fetching, prefer a local fetch method under your control.
Capability Analysis
Type: OpenClaw Skill
Name: jina-web-fetcher
Version: 1.0.0
The skill instructs the AI agent to make external network calls using `curl` to `https://r.jina.ai/` with user-controlled URLs (e.g., `http://目标URL`). This introduces a supply chain risk by relying on an external, third-party service. More critically, it creates a vulnerability for Server-Side Request Forgery (SSRF) against the `r.jina.ai` service if a malicious user provides an internal IP or sensitive URL. Additionally, the direct embedding of user input into a shell command in `SKILL.md` poses a prompt injection risk for the agent if it does not properly sanitize the user-provided URL before execution.
Capability Assessment
Purpose & Capability
The name/description claim to fetch arbitrary webpages (including Google results) via Jina AI. The SKILL.md shows exactly that: curl to r.jina.ai/http://<target>. That is coherent with the stated purpose, though the explicit claim of 'bypass search engine restrictions' raises legal/ToS concerns rather than a technical incoherence.
Instruction Scope
Instructions directly tell the agent to proxy arbitrary URLs through https://r.jina.ai. This means any requested URL (including internal/private addresses or pages requiring auth) and the fetched page contents will be sent to a third party. The SKILL.md does not warn about this, nor does it constrain scope (only public URLs), so the runtime behavior can collect and transmit sensitive data.
Install Mechanism
This is an instruction-only skill with no install steps or code files, so there is no on-disk install risk.
Credentials
The skill requests no credentials, but its ability to fetch arbitrary URLs via a remote proxy is disproportionate: it can be used to access internal network endpoints (SSRF) or to exfiltrate private content to r.jina.ai. No environment variables are required or declared, and the SKILL.md does not disclose that data will be sent to a third party.
Persistence & Privilege
No persistent install or 'always' privilege is requested; the skill is user-invocable and uses normal autonomous invocation defaults.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install jina-web-fetcher - After installation, invoke the skill by name or use
/jina-web-fetcher - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
Jina Web Fetcher 1.0.0 — Initial release
- Fetch web content using Jina AI by proxying any URL.
- Supports scraping Google search results.
- Simple usage via curl commands; no installation required.
- Provides example commands for popular sites.
Metadata
Frequently Asked Questions
What is Jina Web Fetcher - 网页抓取?
使用 Jina AI 抓取网页内容，绕过搜索引擎限制。支持任意URL，支持 Google 搜索结果抓取。 It is an AI Agent Skill for Claude Code / OpenClaw, with 1669 downloads so far.
How do I install Jina Web Fetcher - 网页抓取?
Run "/install jina-web-fetcher" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Jina Web Fetcher - 网页抓取 free?
Yes, Jina Web Fetcher - 网页抓取 is completely free (open-source). You can download, install and use it at no cost.
Which platforms does Jina Web Fetcher - 网页抓取 support?
Jina Web Fetcher - 网页抓取 is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Jina Web Fetcher - 网页抓取?
It is built and maintained by xaiohuangningde (@xaiohuangningde); the current version is v1.0.0.
More Skills