Reforge Build

Overview

Reforge Build is a prototyping tool designed to generate and re-theme high-fidelity UI concepts from structured context, design inputs, and data. Rather than starting from a blank canvas, teams can combine artifacts like functional specs, wireframes, prompts, and JSON datasets to quickly produce polished product mockups. In recent examples, Reforge Build has been used to turn Figma wireframes and enriched data files into modular interfaces that can be re-skinned or repurposed simply by swapping the underlying data.

For AI Product Managers, Reforge Build matters because it compresses the path from idea to testable artifact. It supports rapid exploration of product directions, helps teams validate messaging and UX assumptions earlier, and makes it easier to generate realistic prototypes for internal reviews or user research. Its emerging workflow also reflects a broader AI product pattern: context engineering, where the quality of outputs depends heavily on how well specs, design structure, and data are assembled.

Key Developments

• 2026-01-09 — Brian Balfour introduced the Variations feature in Reforge Build, which auto-generates four distinct mockup directions to speed up product discovery and concept exploration.
• 2026-02-11 — Cole Hoffer shipped messaging agents on Reforge Build: AI bots that open real browsers, assume distinct user personas, and run parallel end-to-end flows to surface feedback on messaging clarity, value proposition gaps, and conversion drivers.
• 2026-05-04 — Reforge Build was used in a context-engineering workflow to create a polished music genre detail page from a Figma wireframe, functional requirements, and a JSON data file enriched with Claude and a custom MCP server. The prototype could be instantly re-themed by replacing the `data.json` file.

Relevance to AI PMs

1. Rapid prototype generation from structured inputs AI PMs can move from PRD-level thinking to realistic UI prototypes faster by combining specs, wireframes, and structured data. This is especially useful for testing new product concepts before committing engineering resources.

2. Better discovery through fast variation and iteration
Features like Variations make it practical to explore multiple directions at once instead of debating a single design path. PMs can use this to compare interaction patterns, positioning, and presentation styles during early-stage discovery.

3. Tighter feedback loops for messaging and UX
With messaging agents and browser-based evaluation flows, AI PMs can assess how different personas might interpret landing pages or product experiences. This helps identify clarity issues, conversion blockers, and value-prop weaknesses much earlier.

Related

• cole-hoffer — Associated with shipping messaging agents on Reforge Build for automated product and messaging feedback.
• messaging-agents — A notable Reforge Build capability that uses AI personas in real browser sessions to evaluate flows and messaging.
• brian-balfour — Highlighted the Variations feature, connecting Reforge Build to faster product discovery workflows.
• variations — A Reforge Build feature that generates multiple mockup directions automatically.
• figma — Used as an upstream wireframing input for Reforge Build in a structured prototyping workflow.
• claude — Used to generate and enrich the JSON data that powered a modular Reforge Build prototype.
• context-engineering — Closely tied to how Reforge Build is used effectively: assembling layered context from specs, design, and data to improve output quality.