Find Broken Links and Understand Redirects in LLMIC
Use LLMIC to inspect broken-link targets, source pages and redirect evidence. Export a useful repair handoff and verify the published change with a fresh crawl.
Open the Broken Links dashboard after a crawl and choose one destination that matters to a reader. Your goal is to answer three questions: what failed, where does it appear, and what should the owner do next? A long list of URLs is less useful than one repairable record with its source intact.
This guide describes the development workflow checked on 1 October 2026. The website currently offers LLMIC 1.0.5; some controls described here may differ from that installer. Review the download page and release notes, or confirm a required control with support before subscribing. This guide is not a claim that a development build has been released.
Real Product Evidence: One Repair, Checked Again
We ran a controlled two-page test in LLMIC 1.0.5 (build 6) on 3 October 2026. Both pages linked to the same missing worksheet. The first scan found one confirmed 404 affecting two source pages. We edited both source links, then crawled the same two URLs again. The fresh scan found zero broken links.
See the real before-and-after scans
The repair was made in the website source: both links to missing-worksheet.pdf were changed to a working setup-worksheet.html page, and the download wording was changed to “Open the setup worksheet”. LLMIC identified and rechecked the issue; it did not automatically edit the website.
Scope: URL List with exactly two source pages; internal links only; external links and resources excluded. Both runs checked 3 unique targets across 4 link occurrences, with source and rendered coverage of 2/2. Compare the 15:52 and 15:56 IST runs only; the earlier point visible in the history chart used a different scope. This is a controlled product demonstration, not a customer result or a claim about rankings, traffic or revenue.
The repaired demo page and help page are kept out of search indexing.
Prepare a Crawl That Covers the Sources
Run or load a crawl containing the pages you want to inspect. Check the host, capture date, exclusions and whether JavaScript rendering was used where required. Broken Links works from captured page references, so a page outside the crawl cannot contribute a measured source relationship to that report.
If a link is added after scripts run, source HTML alone may not contain it. Keep rendered coverage visible. An incomplete crawl can still help you investigate a known problem, but its counts should be described as the observed set, not as a complete inventory of the website.
Choose Crawl References for Discovery
In Broken Links, use Crawl references to analyze the links and resources captured in the active crawl. Open Scan options to review whether external links and resources are included. Resources include images, scripts, stylesheets and iframes; they can require a different repair owner from ordinary article links.
Choose Scan links, or Re-scan when reviewing an existing report. Wait for the result and check its freshness and coverage before interpreting the headline totals. If a report belongs to an older crawl, do not combine its numbers with a new source set as though they were one observation.
Read the Status and Scope Together
The dashboard separates missing resources, server errors, redirects and uncertain checks. Use Internal or External scope to narrow the work. A link to another domain may reject automated access or become rate limited; that observation needs review rather than a reflex decision to delete the link.
Start with an important navigation, setup or download journey. The number of affected pages helps identify repeated problems, but it does not replace editorial judgment. A single failed installation link may deserve attention before a larger set of low-value historical references.
Open the Destination to Find Its Sources
Select a target row to inspect the pages that contain it. Review the source address, anchor text, element type and the recorded source or rendered layer. Open a source page to understand the promise surrounding the link. This is the step that turns “a PDF failed” into “replace the worksheet link in this guide.”
An empty source list has more than one explanation. The address may have come from an import or another discovery route, or the crawl may not have captured the necessary document layer. Read the evidence state and coverage before claiming that no page links to it. Keep uncertainty attached to the record.
Validate an Existing List Without Inventing Attribution
If you already have candidate destinations, use Validate imported URLs. Paste one target per line or choose the CSV or TXT import control. Review the imported values before running Locate. A spreadsheet can contain spaces, unexpected columns or addresses from a different host; correct the input rather than interpreting malformed values as website defects.
The workflow checks the destinations and looks for measured references in the current crawl. Imported membership does not prove an internal link exists. When there is no captured source, record that state and decide whether a wider or rendered crawl is needed. Do not create an editorial ticket naming an unobserved page.
Export a Handoff With the Context Intact
Use the export control to save CSV or XLSX when the report is available. Review the active filters first: an export intended for the internal-links editor should not silently omit the internal records the task needs. Keep the report date and crawl scope with the file you share.
The CSV implementation preserves fields such as status, target URL, source page, anchor, document layer and evidence state. A destination can occupy several rows because it appears in several places. That repetition is useful context, not necessarily a duplicate to delete. Add the owner and approved action in your task system.
Review the Fix Before Publishing It
LLMIC identifies evidence and supports organizing reviewed work. It does not give a missing URL the knowledge to choose the correct replacement. Ask the content owner whether to restore, replace, redirect or retire the destination. Inspect replacement citations and downloads for meaning, not only successful responses.
Where the development build offers Add source-page fixes to the Action Queue, review those tasks before using them. Creating or completing a task is a workflow action. It does not establish that your website changed. Publish the approved edit through your normal website process and keep the acceptance criteria attached.
Recheck the Original Journey
Run a fresh compatible crawl after the change, then inspect the source and destination again. Check repeated template occurrences as well as one convenient example. A fresh successful target alone does not prove that the old source link was updated or that the new file is the intended version.
Record verified, still failing and unable-to-check outcomes separately. A request blocked during the retest stays unresolved. Compare counts only when the scopes are comparable, and report search performance separately from technical repair. LLMIC cannot guarantee that a repaired link will produce more traffic or a subscription.
| What You Need | Where to Look | What It Establishes |
|---|---|---|
| Failed target | Destination row | The request observation |
| Page containing the link | Expanded source references | A measured source relationship |
| Anchor and element | Occurrence details | What the reader or resource expected |
| Uncertainty | Evidence and coverage information | What still needs checking |
| Handoff file | Export CSV or XLSX | Saved observations for review |
Frequently Asked Questions
Why Does LLMIC Ask Me to Run a Crawl First?
The dashboard needs captured page references to show where a destination appears. Start with a crawl that includes the relevant sources. An imported list can identify targets to validate but cannot substitute for measured source relationships.
Which Mode Should I Use to Find New Broken Links?
Use Crawl references for discovery from the active crawl. Use Validate imported URLs when you already have a list of candidate targets. Confirm the actual controls in your installed version; this guide describes the checked development workflow.
Why Does a Target Have No Source Page in the Report?
It may not have a captured reference in the current crawl, or the evidence may be partial. Check the crawl scope and rendered coverage. An imported or sitemap URL is not proof that another page links to it.
Can I Limit the Review to Internal Links?
The development dashboard includes Internal and External scope filters. Set the scope deliberately and note it in your handoff. A filtered count should not be presented as the number of all links across the whole site.
Why Are Several Export Rows Using the Same Target?
The export can preserve one row per source occurrence. A shared target may appear on several pages, with different anchors or element types. Keep those relationships when assigning and verifying the repair.
Does Adding an Action Queue Task Change My Website?
No. It organizes a proposed source-page fix for review. Publish the approved change through your website’s editing or deployment process, then collect fresh evidence. A task marked complete is not proof of a live correction.
Are All These Controls in the Public Download?
Not necessarily. This guide documents the development workflow checked on 1 October 2026. The download page identifies the public release. Confirm your required controls against release notes or support before purchasing for a specific workflow.
Sources, Statistics and Measurement Notes
Technical references checked on 1 October 2026: How HTTP Status Codes Affect Google’s Crawlers (Google); Redirects and Google Search (Google); and SEO Link Best Practices (Google). These explain response behavior, redirects and crawlable links; they do not predict traffic uplift. All worksheet counts and scenarios on this page are illustrative, not customer results. Product labels were checked against the LLMIC development source on 1 October 2026. The public installer may differ; check its release notes before buying for a specific control.
Choose Your Next Step
If you have not yet collected the evidence, follow How to Find Broken Links Across Your Website. If you already know what failed, use the Broken Link Repair Checklist. After publishing the edit, use How to Verify Broken Link Fixes.
LLMIC is a Mac application with paid access and no automatic free trial. Check the public download and release notes before choosing it for a development-build control.