Free Product Feed Image Link Audit

A perfect product photo cannot appear in Shopping or a marketplace when the feed points to an empty, duplicated, insecure, or malformed URL. This free audit inspects the CSV boundary where catalog systems often fail. Paste a feed or load a local CSV and the browser parses up to ten thousand product rows. It requires id and image_link columns, then reports row numbers for missing IDs, empty image links, invalid URLs, non-HTTPS links, duplicate URLs, and file paths that do not look like supported product-image formats. The CSV remains on the device and is not uploaded by this tool.

Product titles, attributes, and URLs can contain commas, quotes, and line endings. Splitting every line on a comma creates false columns and misleading errors. The audit uses a bounded parser that understands quoted fields and escaped quotes. It stops on an unclosed quoted field, files larger than the local limit, or more than ten thousand product rows instead of consuming memory indefinitely. This is a preflight tool, not a full feed-management platform. It focuses on image-link integrity so the result remains fast, understandable, and directly connected to a paid repair workflow.

A missing ID prevents reliable mapping between the row, SKU, approval, and repaired asset. An empty image_link provides no primary image. An invalid URL cannot be parsed consistently. HTTP links are insecure and may be rejected, redirected, or blocked. A duplicate may be intentional for variants, but it deserves review because it can also indicate a copy error. An unsupported extension warning means the path does not visibly end in JPEG, PNG, WebP, or GIF; dynamic image services can still return a valid content type, so verify the response rather than automatically deleting the row.

The browser does not request every remote URL. That protects privacy, avoids cross-origin failures, and prevents the tool from becoming an uncontrolled crawler, but it also means the audit cannot confirm HTTP status, redirect chains, robots access, MIME type, byte size, pixel dimensions, freshness, or whether the image matches the product. After structural errors are fixed, sample actual URLs through an authorized server-side validator or platform diagnostics. Use the Google Shopping Product Image Checker on downloaded representative files to inspect objective pixel and size rules. URL integrity and image-content integrity are separate layers.

Export a copy of the current feed and keep it immutable. Run the audit, group issues by code, and repair deterministic failures first. Restore missing IDs from the system of record, generate or select approved primary images, move assets to stable HTTPS URLs, and investigate duplicates by variant. Do not globally replace strings without checking how the commerce platform constructs URLs. Validate a small canary feed, submit it, and compare platform diagnostics before updating the full catalog. Keep the old feed and a mapping from old to new URL so rollback remains possible.

Use stable, descriptive storage keys that include SKU or another durable identifier, channel when necessary, and a version or content hash. Avoid spaces, credentials, temporary signed URLs, and query parameters that expire before the platform crawls them. When an image changes materially, use a new versioned URL or follow the platform’s documented recrawl process; relying on the same cached URL can delay updates. Do not encode personal information or secret account data in paths. The delivery manifest should preserve checksum and byte size so teams can prove which binary a feed URL referenced at approval time.

A useful audit ends with actions, not a spreadsheet of red cells. Map each failed row to a workspace and SKU, attach the approved source image, run the necessary background or main-image repair, and route the output to review. Once approved, produce a Merchant or Shopify feed with the new URL and a ZIP containing the manifest. Yito Campaign delivery records include item, channel, locale, model, cost, version, checksum, and approval time. The organization can share generation credits while keeping clients, brands, comments, and deliverables separated.