A few things before
your first review.

RUDU adds a visual feedback layer to the website you are reviewing. Pin comments, select elements, type or record feedback, and keep screenshots and page context with your notes. Export the review to give your coding tool a clearer implementation brief.

RUDU is built for people improving websites with AI coding tools, especially solo builders and small teams. It also helps designers and product owners capture a clear round of feedback for the person implementing it.

The current product targets a Chrome or Edge extension with a Windows local companion. The extension provides the review interface; the companion supports local storage, transcription and export. Supported release versions and installation instructions will be published with beta access.

Your coding agent carries out the implementation you request. With a supported MCP connection, it can read your RUDU review and report outcomes back to the notes. You review the updated site and decide what needs another pass. Connections and update support depend on the agent and integration.

A supported MCP connection lets your agent read review notes and their context. Ask the agent to work through the notes you choose and, where the integration supports it, report outcomes back to RUDU. You can also download Markdown, structured context and images for a portable handoff. Supported connections and setup details will be published with the beta release.

The review package contains REVIEW.md, AGENT_CONTEXT.md, session.json and relative screenshot assets, with the available page, target and viewport context. Your original feedback stays useful outside RUDU. You can download an illustrative sample from the How it works page.

Where the MCP integration supports reporting outcomes, your agent can record what it actioned and what remains partial, blocked or in need of clarification. Those reported outcomes update the relevant notes in RUDU. You can then inspect the actual site and continue the review; an agent report does not replace your judgement of the result.

Starting or resuming an agent task from the RUDU interface is planned. In a supported connected workflow, ask the agent to begin from its own interface. MCP provides access to review context and supported updates; connecting it does not automatically start work.

Yes. Voice notes use the local companion for transcription. Check and edit the transcript before using it in your handoff. Typed notes are available too.

Review desktop, tablet and mobile viewport profiles and retain that context with each note. These profiles help inspect responsive layouts; they do not reproduce every characteristic of a physical device.

Local-only reviews are not uploaded to RUDU Cloud. Capture, storage and voice transcription use the local workflow. If you share an export or use an AI connection, that tool receives the context you provide under its own policies.

Local capture and export are designed to work without a RUDU Cloud account. Connected features require an account and the appropriate setup. The public cloud connection reads reviews that you choose to sync.

Local capture is designed to continue when RUDU Cloud is unavailable. Connected reviews retain pending sync state for later delivery. Your chosen AI tool may have its own connection requirements.

Pricing and beta terms have not been announced. We will share them when confirmed; there is no paid signup or checkout on this website.

The beta signup list is not open yet. The early access page will be updated when signups are available, alongside details about access timing and the supported release.