Link and Image Tools

Link and image tools operate at two levels: list the records that match a condition, then retrieve the source pages behind an important destination or asset.

Link tools

list_links

Lists link records for a known crawl. Use structured filters to isolate internal, external, failed, redirected, or target-specific links.

get_link_sources

Returns pages that link to a target. Use it to distinguish a one-off content link from a repeated template or navigation issue.

Find internal links with failed destination responses.
Group by target URL. For the ten targets with the most occurrences,
use get_link_sources and return representative source pages.

Useful link fields include URL or target values, domain, status information, internal or external classification, and source context supported by the dataset.

Image tools

list_images

Lists image records for a known crawl. Filter by response status, alt-text state, response size, response time, format, URL, or other allowlisted image fields.

get_image_sources

Returns pages where an image appears. Use it to prioritize shared assets and understand whether missing alt text needs page-specific context.

Find images with missing alt text and high source-page counts.
For each shared asset, show representative source pages and explain
whether the image appears meaningful or decorative based on available context.

Source pages change priority

An asset or destination can appear once, on several important pages, or across an entire template. Source data helps answer:

• Is the fix local or shared?
• Which team owns the source?
• How many user journeys are affected?
• Can one template change resolve the issue?

Interpretation limits

Crawl records show observed technical facts. They do not always reveal editorial intent. An empty alt attribute can be correct for a decorative image, and an external destination can fail temporarily. Include these uncertainties in the final recommendation.