← 返回 Skills 市场
1171
总下载
1
收藏
13
当前安装
2
版本数
在 OpenClaw 中安装
/install jina-ai-reader
功能描述
Fetch clean, AI-friendly Markdown content from any URL using Jina.ai Reader. Bypasses paywalls, handles Twitter/X posts, renders JavaScript-heavy pages, retu...
使用说明 (SKILL.md)
Jina.ai Reader
Fetch clean, AI-friendly Markdown content from any URL. No API key required.
Features
- ✅ Bypasses paywalls (tested with Every.to, Medium, etc.)
- ✅ Handles Twitter/X posts and threads
- ✅ Renders JavaScript-heavy pages (optional wait)
- ✅ Returns clean Markdown format
- ✅ Free, no API key needed
Basic Usage
node {baseDir}/scripts/jina-reader.mjs "https://example.com/article"
Options
| Option | Description |
|---|---|
--wait-ms N |
Wait N milliseconds for JavaScript to render |
--with-images |
Include image captions in output |
--with-links |
Include all links in output |
Examples
# Basic fetch
node {baseDir}/scripts/jina-reader.mjs "https://every.to/article"
# Twitter/X post
node {baseDir}/scripts/jina-reader.mjs "https://twitter.com/user/status/123456"
# Wait for JavaScript rendering
node {baseDir}/scripts/jina-reader.mjs "https://spa-site.com/page" --wait-ms 5000
# With images and links
node {baseDir}/scripts/jina-reader.mjs "https://blog.example.com/post" --with-images --with-links
When to Use
- Search + Read: After Tavily/desearch finds URLs, use this to read the actual content
- Twitter/X: Most tools can't handle Twitter, this one can
- Paywalled sites: Works on many sites that require login
- SPA/JS-heavy pages: Use
--wait-msto let scripts run
Notes
- No API key required
- Rate limits may apply for heavy usage
- For pagination, you may need to fetch multiple pages manually
安全使用建议
This skill legitimately uses Jina.ai's public r.jina.ai endpoint to fetch rendered Markdown, so its behavior is coherent. Before installing, consider: (1) privacy—any page you fetch (including paywalled or private content) will be processed by a third party (r.jina.ai); avoid sending sensitive pages or credentials; (2) legality/terms—bypassing paywalls may violate site terms of service or copyright; (3) runtime—ensure your Node version supports global fetch (Node 18+ recommended). If you need local-only scraping or to avoid sending content to an external service, prefer a local headless-browser-based extractor instead.
功能分析
Type: OpenClaw Skill
Name: jina-ai-reader
Version: 1.0.1
The skill bundle is benign. It provides a straightforward implementation for fetching web content using the Jina.ai Reader service. The `SKILL.md` file clearly outlines the skill's purpose and usage, without any evidence of prompt injection attempts against the AI agent. The `scripts/jina-reader.mjs` script correctly parses arguments and makes a standard HTTP request to `https://r.jina.ai/` to retrieve content, without any signs of data exfiltration, malicious execution, persistence mechanisms, or obfuscation. The capability to fetch arbitrary URLs is inherent to the skill's stated purpose and does not indicate malicious intent or a vulnerability within the skill's own code.
能力评估
Purpose & Capability
Name/description promise (fetch AI-friendly Markdown, handle JS pages/Twitter, no API key) matches the provided code and required binary (node). The script constructs requests to r.jina.ai and prints the returned Markdown — the requested binary and behavior are proportionate to the stated purpose.
Instruction Scope
The SKILL.md only instructs running the included Node script; it does not read local files or request extra env vars. Important privacy note: the script sends the target URL to a third-party service (https://r.jina.ai/), so the page contents (including paywalled or sensitive pages) will be fetched/processed remotely. That is expected for this skill but is a data-exfiltration/privacy consideration users should understand.
Install Mechanism
There is no install spec (instruction-only plus a small included script). No downloads or archive extraction are performed, so nothing arbitrary is written to disk beyond running the provided script with the existing node runtime.
Credentials
No environment variables, credentials, or config paths are requested. The only required runtime is node, which is appropriate for the provided .mjs script.
Persistence & Privilege
The skill does not request permanent presence (always:false), does not modify other skills or system-wide settings, and does not store credentials — no elevated persistence or cross-skill access is requested.
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 jina-ai-reader - 安装完成后，直接呼叫该 Skill 的名称或使用
/jina-ai-reader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
v1.0.1
Update description.
v1.0.0
- Initial release of jina-reader skill.
- Fetches clean, AI-friendly Markdown from any URL with no API key required.
- Bypasses many paywalls (e.g., Every.to, Medium), handles Twitter/X posts and JavaScript-heavy pages.
- Offers options for waiting for JS rendering and including images or links in the output.
- Returns structured content with titles and metadata, suitable for search-and-read workflows.
- Designed to work with a variety of web content types, including paywalled and SPA pages.
元数据
常见问题
jina-ai-reader 是什么？
Fetch clean, AI-friendly Markdown content from any URL using Jina.ai Reader. Bypasses paywalls, handles Twitter/X posts, renders JavaScript-heavy pages, retu... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 1171 次。
如何安装 jina-ai-reader？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install jina-ai-reader」即可一键安装，无需额外配置。
jina-ai-reader 是免费的吗？
是的，jina-ai-reader 完全免费（开源免费），可自由下载、安装和使用。
jina-ai-reader 支持哪些平台？
jina-ai-reader 跨平台运行，可在任意部署了 OpenClaw / Claude Code 的环境中使用（cross-platform）。
谁开发了 jina-ai-reader？
由 soundi（@jiangtianjiao）开发并维护，当前版本 v1.0.1。
推荐 Skills