• ERROR LOGS
  • TIMELINE
  • INDEX
  • HUMAN

How to identify broken archived links with your AI

Connect the Internet Archive Wayback MCP to your agent to audit historical web snapshots. It turns massive archive data into a list of specific failure points.

Ask AI about this page

Short answer

How can I find when my archived pages stopped working?

Your AI uses the Wayback Machine's index to filter historical captures by their HTTP status codes. You'll get a clear timeline of when specific URLs transitioned from working to returning errors like 404s or 500s.

Where every issue lands.

The audit outcomes

The AI processes the archive index to deliver these specific results.

ERROR LOGS

Status code lists

The AI pulls captures that returned specific error codes. You get a direct list of failed requests instead of browsing thousands of successful ones.

TIMELINE

Failure dates

By looking at the capture history, your agent identifies the exact timestamp a page first returned a non-200 status. This helps you pinpoint when content disappeared.

INDEX

URL snapshots

The AI retrieves the full list of available captures for a domain. This provides the raw data needed to check every version of a page.

HUMAN

Manual verification

The AI flags the broken timestamps, but you decide which pages need content restoration or redirects. You handle the actual site fixes.

The workflow

What your AI does when the request arrives.

The AI moves from broad index searches to specific error filtering.

  1. Fetch capture history

    The agent pulls the complete list of historical snapshots for your target domain or URL.

    get_cdx_captures

  2. Filter by status

    The AI narrows down the results to only show captures that returned specific HTTP error codes.

    get_captures_by_status

  3. Identify failure points

    The agent compares successful captures against error captures to find the moment a link broke.

    get_cdx_captures

  4. Summarize findings

    Your agent organizes the error timestamps into a readable report for your review.

    get_captures_by_status

Try it

Copy these to start.

Use these prompts to begin your audit.

Starting points

Replace the domain or URL with the actual site you want to audit.

Internet Archive Wayback Machine Connector

You're all set. Choose your MCP client and follow the setup instructions.

Connector linkhttps://edge.vinkius.com/vk_preview_hnm6MKLSX94NqzQjsRaRbJo66cdvIxnhQfEOtu5s/mcp

Claude Desktop

Follow the steps below to connect in seconds.

  1. 1In Claude Desktop, open Settings → Connectors.
  2. 2Click “Add custom connector” and paste the connector link above as the remote MCP server URL.
  3. 3Click Add and start a new chat — Internet Archive Wayback Machine capabilities are ready to use.
Configuration · claude_desktop_config.jsonCopy
{
  "mcpServers": {
    "internet-archive-wayback-mcp": {
      "url": "https://edge.vinkius.com/vk_preview_hnm6MKLSX94NqzQjsRaRbJo66cdvIxnhQfEOtu5s/mcp"
    }
  }
}
Copy into chat04

  • Find all captures for example.com that returned a 404 error.

  • When did the page at https://example.com/old-page first start returning a 500 error in the archive?

  • List all the historical captures for mydomain.org using the CDX index.

  • Check the Wayback Machine for any captures of mydomain.com that have a status code other than 200.

  • Claude
  • ChatGPT
  • Cursor
  • VS Code
  • Windsurf
  • Claude Code
  • JetBrains
  • Cline

Start here

Connect Internet Archive Wayback once, then ask.

Just click the link to connect the MCP to your client. Your credentials stay encrypted, and you can start asking questions immediately.

Connect Internet Archive Wayback to your AI

FAQ

How this task behaves

  • 01

    Can the AI fix the broken links on my actual website?

    No. The AI only reads historical data from the Internet Archive. You must make any changes to your live site yourself.

  • 02

    Does the AI search the live web or the archive?

    It searches the Internet Archive's snapshots. It does not crawl the live, current version of your website.

  • 03

    Can I find links that were never archived?

    No. If the Wayback Machine never captured a specific version of a page, the AI cannot find it.

  • 04

    Can the AI delete old captures from the Wayback Machine?

    No. The AI only has read access to the archive index. It cannot modify or delete any data in the Internet Archive.

  • 05

    How does it know which status code to look for?

    You tell the AI which code you are interested in, such as 404 or 503, and it filters the captures accordingly.

  • More questions about Internet Archive Wayback? See everything the Internet Archive Wayback Connector can do

Connect Internet Archive Wayback Machine to Claude, Cursor, ChatGPT & more