Imbricalt Viewer

Alt text used to be locked in a CMS. Now it lives inside the image file. The Imbricalt Viewer Chrome extension scans any webpage, detects images processed by Imbricalt, and displays their alt text as a floating label — no backend, no API, no configuration.

How It Works

Imbricalt is an AI alt text generator that embeds descriptions directly into JPEG and PNG image metadata (IPTC/XMP fields). Each processed image carries an imbricalt:processed signature inside the file itself. The Imbricalt Viewer extension reads this signature in two stages:

Stage 1 — Metadata scan. When you visit a page, the extension fetches each JPEG and PNG image and parses its embedded XMP metadata. If the imbricalt:processed signature is found, the alt text is extracted from the dc:description or Iptc4xmpCore:AltTextAccessibility field and displayed as a floating overlay near the image.

Stage 2 — Perceptual hash (Fallback). If the metadata has been stripped (for example, by image-optimisation pipelines, CDNs, or social media platforms), the extension computes a perceptual hash (pHash) of the image using the same DCT-based algorithm that runs on Imbricalt servers. This 64-bit hash is sent to the Imbricalt API, which finds the closest match via Hamming distance. Each image ever processed through Imbricalt is indexed and findable through this method.

Key Features

- Zero configuration. Install the extension. That's it. It runs automatically on every page.
- Works everywhere. Any site serving JPEG or PNG images processed by Imbricalt. WordPress, Notion, Shopify, GitHub, blogs, social media, email — if the image went through Imbricalt, the extension surfaces the description.
- Privacy-first. The extension reads image bytes — never page content, cookies, localStorage, or personal data. The perceptual hash fallback sends only a 64-bit hash (16 hex characters) to the API. No identifying information is transmitted.
- Copy button. Every overlay includes a copy button to grab the alt text instantly.
- Retroactive coverage. Every image ever processed by Imbricalt since launch is indexed. Install the extension and it immediately works across your entire browsing history.
- Non-intrusive overlay. The alt text label appears at the bottom-right of the image as a small, semi-transparent dark panel with amber accents. It stays out of the way while remaining readable.
- Works with all Imbricalt plans. Images processed on the Try (5 free), Free (50/mo), or Pro ($9/mo unlimited) tier are all detected identically.

What This Extension Does NOT Do

- It does not read page text, HTML content, or form data.
- It does not send image content to external servers (the pHash lookup sends only 16 hex characters).
- It does not log or store browsing history.
- It does not require an account or API key.
- It does not inject ads or tracking.
- It does not modify images or page content — only adds a visual overlay.

Permissions Explained

The extension requests no special permissions. It runs as a content script on all URLs (<all_urls>) to scan images on any webpage. The only network request is to the Imbricalt API (imbricalt.synov8studio.com/api/metadata) — a minimal lookup using a 64-bit hash derived from the image bytes.

Privacy

Imbricalt Viewer is built with privacy as a first principle:

- All image processing happens in your browser
- The 64-bit hash sent to the API is lossy — it cannot be reversed to reconstruct image content
- The API returns only the alt text associated with the closest hash match
- No analytics, no telemetry, no cookies

Who Is This For?

- Developers who want alt text to travel with image files across platforms
- Content creators who generate alt text with Imbricalt and want it to surface on any site
- Accessibility professionals verifying alt text on live sites
- Anyone who downloads Imbricalt-processed images and uploads them to different platforms

About Imbricalt

Imbricalt (imbricalt.synov8studio.com) is an AI alt text generator that embeds descriptions directly into image file metadata. Instead of storing alt text in a database or HTML attribute — both of which are lost when the file moves — Imbricalt writes the description into IPTC and XMP fields inside the image itself. The Chrome extension is the read-side of this system, surfacing embedded descriptions wherever the image appears on the web.