Web Clipper

What This Skill Does

The Web Clipper is a powerful utility designed for OpenClaw agents to harvest high-quality textual information from the live web. By leveraging modern parsing engines like Mozilla's Readability, it strips away the noise—ads, navigation bars, sidebars, and trackers—that plague contemporary websites. It then converts the resulting clean HTML into structured, readable Markdown files. These clippings are automatically deposited into your memory/clippings/ directory, complete with YAML frontmatter containing metadata such as the original URL, document title, and publication timestamps, ensuring your research remains organized and traceable.

Installation

To integrate this skill into your agent, ensure you have the OpenClaw environment initialized. Execute the following command in your terminal:

clawhub install openclaw/skills/skills/autogame-17/web-clipper

Once installed, navigate to the skill's directory and run npm install to resolve all necessary dependencies, including axios for network fetching, jsdom for DOM emulation, and turndown for the final Markdown transformation. No further configuration is required to begin clipping.

Use Cases

This skill is indispensable for knowledge workers and researchers. Use it to:

• Archive technical documentation that might disappear or change.
• Collect long-form articles or blog posts for offline analysis or local vectorization.
• Transform complex web pages into simplified study notes.
• Maintain a local, searchable repository of internet-based information that the agent can read later without network constraints.

Example Prompts

1. "Clip the article at https://example.com/deep-learning-guide and save it so I can summarize the main points later."
2. "Go to this URL: https://docs.openclaw.ai/quickstart and save it to my clippings folder for offline reference."
3. "Find the documentation for the new API at https://api.example.com/docs and clip it so the agent can reference it during our session."

Tips & Limitations

While the Web Clipper is robust, it is designed for content-heavy pages (blogs, articles, docs). It may struggle with highly dynamic single-page applications (SPAs) that require heavy JavaScript execution to render content. Ensure the URL is accessible to your agent's network environment. For best results, use canonical URLs rather than redirected shortlinks. Keep in mind that as the tool saves files to your local disk, regular housecleaning of the memory/clippings/ folder is recommended to prevent local storage bloat.