Anyone who has used an AI Agent for content or data collection knows that scraping websites is the dirtiest and most exhausting task.

In my usual workflow, I rely on Chrome cookies to fetch data; when cookies expire, they become useless.
When encountering sites with strict Cloudflare or anti-scraping measures, I get a 403 error directly.
And for content on X, forget it—login sessions often expire, and once API quotas are used up, I have to switch to another solution.
For each link, I have to prepare three or four fallback layers, and often, even after reaching the last layer, I still can't get the data.
All the effort spent on "getting the data in" exceeds the effort of "using the data."

I tried XCrawl and installed its skill on my OpenClaw bot.
First test—telling the bot "fetch the content of "—it returned structured markdown with dozens of prediction markets' odds, trading volume, and deadlines.
For pages rendered dynamically with JS, one request does the job.

Second test—more aggressive—I fed in a link to my own X Article.
It returned thousands of words along with views, likes, and bookmarks—all in one go.
X content is notoriously hard to scrape; previously, I had to write a separate logic just for that. Now, it's just one sentence.

Looking at the consumption, each request costs 1-2 credits.
Built-in residential proxies and JS rendering mean I don't need to set up infrastructure myself.
The output markdown can be directly fed into an LLM or stored in a database—no secondary cleaning needed.

There are five API modes—single page scraping, full-site crawling, sitemap, search, SERP—covering most daily collection scenarios.
OpenClaw users can just install a skill to use it; registering gives 1000 credits, enough to run for a while.

Honestly, this layer of data collection infrastructure should have been a service long ago.
Building it myself is too costly, and maintenance is even more exhausting.
On-demand calls save time, allowing me to focus on truly valuable analysis and decision-making.