September 2026

We introduced Wirechunk 2.0 on September 24. Since then, we’ve added more ways to shape your website, let our built-in AI inspect its work in a browser, and made forms more useful for the team receiving the inquiries.

Give your site more of your own style

Wirechunk now supports primary and secondary color palettes with coordinated light and dark versions, plus a choice of neutral colors. Ask the AI to work from your brand color and apply it across your site, or give a particular section, header, or footer its own colors. These styling options are available through agents and the API.

Section backgrounds now extend across the screen while the content keeps its reading width. That makes it easier to distinguish a service introduction, a contact section, or a call to action.

New palette checks flag insufficient contrast between the primary and secondary palettes’ main fill and text colors in both themes, helping you spot hard-to-read combinations as you adjust your branding.

You can also ask your agent to align text to the left, center, or right within its block.

The built-in AI can open a browser

Wirechunk’s built-in AI can now browse rendered websites and draft previews, inspect page controls, and take screenshots. It can click through navigation, interact with forms, and resize the browser to check a layout at different screen sizes.

For example, ask it to look at your contact page on mobile, check the navigation, or show you a screenshot of a revised section. Screenshots can appear directly in the conversation, so you can see what the AI is looking at as you discuss changes.

The browser stays with the chat across replies while its session is active. These browser tools are available to Wirechunk’s built-in AI.

Group your navigation into dropdowns

Header navigation now supports dropdowns. Group related links under labels such as “Services” or “Resources” to give visitors a clearer route through a larger site.

You can create, reorder, and remove dropdowns and their links in the editor. Each dropdown supports up to 18 links, with longer lists arranged into two columns on desktop. On mobile, the same links appear in labeled groups, and long menus scroll within the screen.

Add icons and custom HTML

The new Icon block lets you add visual cues that help visitors scan a list of services or steps. Search the Phosphor catalog in the editor, choose a size, and place an icon alongside the text it supports. Agents can search the same catalog and set colors for light and dark mode.

For exceptional cases that need custom markup, an HTML block lets you supply HTML directly. It also has an optional Markdown field for the page’s Markdown version. Use that field to carry the relevant information through to agents reading Markdown; without it, the HTML block is omitted from that version.

Clearer form setup and CRM mappings

We’ve improved the details around form setup and delivery:

• Clearer CRM mappings. Integration configuration now explicitly describes each provider’s supported fields. Name mapping consistently lets you choose one full-name field or separate first- and last-name fields, making setup clearer for people and agents using the API.
• HighLevel address fields. Map your form’s street address, city, state, postal code, and country fields to the corresponding fields in the HighLevel contact record, saving your team from copying them out of an inquiry.
• Longer labels. Form field labels can now contain up to 400 characters, giving longer questions and checkbox labels more room.

Collect customer ratings and feedback

The new Review Form block collects star ratings and customer feedback in a private Reviews screen, giving your team one place to see how customers felt about their experience and read any concerns they shared.

You configure a Google review link and the minimum rating at which to show it. Ratings meeting that threshold are saved before the visitor sees the invitation to leave a Google review. Below the threshold, the form offers optional name and written-feedback fields before submission. Set the threshold to one star to offer the Google review link for every rating.

The form supports mouse, touch, and keyboard input. Saved ratings and any accompanying feedback are available in the Reviews screen; leaving a review on Google is a separate step the visitor chooses to take.

Branded images for your posts

Wirechunk now generates share cards using your site’s primary palette, site name, and post title, with a neutral fallback when the palette would make the text hard to read.

Blog listings use a post’s featured image, its share image, or a generated card when neither is supplied. Social previews use an explicit share image first, then a featured image, then the generated card. You can publish a post without having to design a separate image just to give its link a useful preview.

The site-wide default share-image setting has been removed in favor of these page and post choices. Page and post titles now have a 120-character limit.

Smaller improvements

• More consistent Markdown. Saving content standardizes common formatting markers, keeping the source consistent as you switch between agent and manual edits. Its structure and rendered appearance stay the same. The visual editor also stops automatically turning typed URLs into links; you can add links intentionally.
• Another discovery path for agents. Published sites now support /llms-full.txt, helping agents that look for this address discover information about your site.
• Feedback through your agent. Report a limitation or problem from the conversation where you encountered it. Your agent can offer to send product feedback to Wirechunk and submit it with your agreement.

Open your site in the Wirechunk app to try the new blocks and navigation, or ask the AI to help apply these changes to your draft. Review the result in a preview and publish when you’re ready.