Rezonant launches screen-and-voice PRD capture for Claude Code and Cursor

• Rezonant is pitching a new layer between product intent and coding agents: according to hasantoxr's launch thread, you record your screen, narrate the feature, and get a structured PRD that can feed Claude Code or Cursor.
• The setup starts with context connectors. In hasantoxr's integrations post, Rezonant lists GitHub, Linear, Jira, Figma, Notion, Granola, and Confluence as sources for grounding the brief.
• The capture flow runs through a Chrome extension called Rezonant Alter, which hasantoxr's extension demo says can record a live product, prototype, or Figma file while you talk through changes.
• The output is not just a document. As hasantoxr's agent handoff post describes it, the generated spec can turn into engineering tickets or go straight into Rezonant's own coding agent.
• Access looked early and promotional on day one: hasantoxr's signup post said signups were open with 3,000 credits instead of 1,000, and that the Chrome extension was already live.

You can jump straight to Rezonant's signup page, watch the launch demo, and skim the individual posts on integrations, voice-to-spec, and the multiplayer workspace. The weirdly useful part is the input format itself: instead of typing a long prompt, the extension walkthrough has you point at the UI and narrate like you're already in a scoping meeting.

The capture surface is a Chrome extension called Rezonant Alter. In hasantoxr's extension walkthrough, the input is a live product, prototype, or Figma file plus a spoken walkthrough.

That puts screen context, references, and narration into the same step. For creative and product teams, the notable shift is that the raw material is a review session, not a blank spec template.

According to hasantoxr's voice-to-spec post, Rezonant combines what was said out loud with codebase context and outputs a structured brief with edge cases and acceptance criteria.

The workflow in the thread breaks into three concrete stages:

• Connect existing tools for context, per hasantoxr's integrations post
• Record a walkthrough in the browser, per hasantoxr's extension walkthrough
• Generate the brief from the recording, per hasantoxr's voice-to-spec post

Rezonant's claim is that the brief does not stop at documentation. In hasantoxr's agent handoff post, the output becomes engineering-ready tasks for Linear or Jira, or goes directly to Rezonant's coding agent.

The launch framing in hasantoxr's problem statement is that coding agents sped up implementation, but not the handoff that decides what gets built. Rezonant is trying to make that handoff machine-readable from the start.

The collaboration layer is a multiplayer cloud workspace. hasantoxr's workspace post says PMs, engineers, designers, and founders work from the same source of truth while specs stay current as the codebase changes.

A separate day-one example in hasantoxr's founder workflow post pushes the idea even further: point at the product, describe the change out loud, and Rezonant opens a PR. That is a different claim from "better docs," and it is the freshest, most ambitious part of the launch thread.

The launch also came with an early access style offer. In hasantoxr's signup post, hasantoxr said new users could sign up with 3,000 credits instead of 1,000, and linked to the signup page.

That same post positioned Rezonant as a coordination layer for teams already shipping with Claude Code, Cursor, or other agents, which makes the product less like a standalone editor and more like workflow glue around the tools teams already use.