Page Checker: Is This Page Really Working?
A link can answer 200 OK and still show an error or an empty page. Get a verdict and the text that triggered each finding. We read the returned HTML without running JavaScript.
60 free checks per IP per UTC hour. No account needed.
Verdict:
What each finding means
- HTTP status or response type
- The link returned something other than a 200 HTML page.
- Near-empty page
- Fewer than 50 visible text characters were found. Scripts, styles and hidden elements do not count.
- Missing page or error text
- The title or short body looks like an error page, even when the server answers 200.
- Home page redirect
- You asked for a page path, but the link sent you to a home page.
- Home page canonical
- The page names the home page as its preferred URL. Check whether the intended page was replaced with a fallback.
- Unfinished copy
- TODO, FIXME, TBD, XXX, placeholder text, lorem ipsum, template braces or a coming-soon message may need review.
Worked example
You check https://example.com/old-article/. It answers 200 OK, but its short body says Page not found
. The verdict is broken, with that matching snippet. Repair the link or return the intended article.
A long, healthy article that mentions a 404 can still get working. A page containing TODO: add the summary
gets suspect so you can review the copy.
How to use this tool
Paste a link. We fetch it like a browser would and tell you if it really shows something.
Frequently asked questions
What does the Page Checker read?
It reads the HTML returned for one public HTTP or HTTPS link and follows up to five redirects. It does not run JavaScript or sign in. A page that needs JavaScript may look empty here; open it in your browser too.
What do working, suspect and broken mean?
Working means none of these fixed rules found a problem. Suspect means unfinished copy, a canonical URL pointing to the home page, or an error-style title over a long page (a real article about 404 errors can look like that). Broken means a non-200 response, non-HTML response, fewer than 50 visible text characters, error text or a redirect from a page path to a home page. These are clues to review, not a guarantee.
Will an article mentioning 404 be flagged?
Error words in the body count only when it has fewer than 1,500 visible text characters. An error in the title still counts. Unfinished-copy markers can be found on pages of any length.
Is the Page Checker free?
Yes. You get 60 web checks per IP per UTC hour without an account. API and MCP calls share 20 free agent calls per IP per UTC day and the global free ceiling. No model or paid API is used.
What are the fetch limits?
Each check has a 10-second deadline and a 2 MB HTML cap. Only public HTTP and HTTPS addresses on ports 80 and 443 are accepted. Each redirect is checked again. Reports include up to 40 findings with snippets of at most 180 characters. URLs and page contents are not stored.
API Tool Calls MCP
Use it from an AI agent
This tool is also an MCP tool named page_check at https://apitoolcalls.com/mcp.
Plain JSON: GET /api/page-check?url=.
Agent tools are free: 20 calls a day per IP, no key needed.
Example tools/call arguments:
{"name":"page_check","arguments":{"url":"https://example.com/article/"}}