chrome-mcp-web-fallback

What problem does it solve?

This Skill addresses web retrieval failures caused by blocking, dynamic rendering, bot protection, redirects, transport errors, or other non-browser HTTP limitations by retrying once via a browser-driven path.

Core Features & Use Cases

• Failure-aware retry flow: Captures a one-line symptom from the initial attempt (HTTP status, timeout/TLS/DNS/connection errors, bot interstitials, or consent/login redirects).
• Scoped chrome-mcp extraction: Re-opens the same target with chrome-mcp, waits for the page to settle, and extracts only what the user’s request needs.
• Traceable dual outcome: Returns both initial_attempt (tool + failure symptom) and fallback_attempt (chrome-mcp + extracted result or failure), then stops to prevent loops.
• Use case: Validate and scrape data from a website where curl/fetch works inconsistently, such as when a page intermittently returns 403/429 or requires client-side rendering.

Quick Start

Use the chrome-mcp-web-fallback skill when a direct web request to your target URL fails due to blocking or dynamic rendering, and ask it to extract the specific data you need from that page.