| name | chrome-fetch |
| description | Fetch web page content using Chrome DevTools MCP when WebFetch fails, returns domain verification errors, or the page requires authentication/JS rendering. Uses the user's running Chrome browser via autoConnect. (agent) |
Chrome Fetch Skill
Fetch web content through the user's Chrome browser when WebFetch cannot (domain restrictions, authentication, SPAs needing JS rendering).
Workflow
new_page(url) to open the target URL in a new tabwait_for(text, timeout: 15000) with an expected keyword — if it times out, proceed anyway- Extract content via
evaluate_script:
() => { const main = document.querySelector('main, article, [role="main"], .content, #content'); return (main || document.body).innerText; }
For long pages, append .substring(0, 50000) to truncate.
- If extraction returns empty (shadow DOM, iframes), fall back to
take_snapshot() and read StaticText entries
close_page(pageId) to clean up (skip if it's the only open page)
Key Details
- Requires chrome-devtools-mcp with
--autoConnect and Chrome 144+
- The user's cookies/sessions are available, so authenticated pages work
- For complex SPAs, you may need to wait longer or interact (click, scroll) before content appears
Troubleshooting
Chrome Connection Issues: If you encounter "Could not connect to Chrome" errors, see the Chrome Setup Guide for detailed troubleshooting steps.
Support
If you encounter any issues with this plugin, please report them following our Support Guide. Your feedback helps improve the community experience!
