Last verified: August 27, 2026
What You Can Now Do
Context Memo now scans a verified domain with an automated vision pass, harvests the images already published on it, and surfaces the full set in one browsable interface called the Assets Landscape Viewer. You no longer upload a brand kit or hunt for screenshots before writing a memo. You see the actual visual inventory a model can find on your site, sorted by asset type, page of origin, and recency, with the source URL attached to each file. The task this changes: instead of asking a designer which logo or product shot is current, you check the viewer and work from what is live on the domain today.
Where It Is in Context Memo
The Assets Landscape Viewer sits in the Assets area of the platform, inside the AI visibility category alongside scanning and memo publishing. Open Assets and you will see the scan trigger for any domain you have verified, and, once a scan completes, the populated grid of harvested images with type, source page, and recency on each card.
How to Use It
- Verify the domain first. Complete verification by email match, DNS TXT record, or meta tag. Only authorized representatives can harvest and manage a brand's images, so the scan will not run against an unverified domain.
- Open Assets and point the scan at the verified domain. The crawler walks the site and the vision pass identifies what each image is. You do not select pages or upload anything.
- Review the populated viewer. Logos, product UI, team headshots, diagrams, and marketing graphics come back in one grid. Each asset carries the source URL where it was found, so you can trace any file back to the page serving it.
- Sort by recency to find the drift. Old screenshots of a redesigned product, retired logo variants, and photography that no longer matches current positioning show up as dated assets on live pages. Fix the page, not the library.
- Work from the harvested set when you write. Use the inventory as the reference for which images belong in memos, landing pages, and structured content that AI systems parse, rather than pulling from a shared drive that may be out of sync with the site.
- Re-scan after a site change. The library reflects the domain as of the last scan. Run a new scan after a redesign, a product launch, or a page migration and the viewer updates to match.
Why We Built It
Marketing teams kept telling us the same thing: nobody can say with confidence which brand images are actually public. The files live in a DAM, a shared drive, and a designer's laptop, and none of those three match what the website is serving. That gap has a cost now that AI answers carry images next to text citations. If the images a model can find are outdated product screenshots, off-brand photography, or third-party thumbnails, the visual story attached to your citation is not the one your team approved, and you never see the substitution happen. Tying the library directly to the live domain removes the sync problem. What is on the site is what is in the viewer.
What It Does Not Do Yet
A few limits worth knowing before you plan around this:
- Refresh is scan-triggered, not continuous. Site changes appear in the viewer on the next scan, not the moment they go live. Build a re-scan into your launch checklist.
- Harvest is limited to on-domain images. The viewer collects what your verified domain publishes. Images hosted by third parties, partners, or review sites are outside the harvest and still need to be handled at the source.
- This is not a DAM replacement. There is no upload path, no version history, and no rights or usage metadata. Dedicated asset management tools do that job with far more depth. Treat the viewer as the authoritative record of what is public, and keep your DAM as the record of what exists.
- Value depends on an existing web presence. Brands with a thin site or no published imagery will get a thin harvest. Publish the pages and the visuals first, then scan.
One more practical note: verification is a hard gate, not a soft one. If you manage multiple brand domains, verify each one before you expect a harvest, because the scan is scoped to the domain you have proven you represent.