Business Driven Development AI Agent

PurplePrint is an AI product-design coach that turns ideas into product decisions, design documents, and build-ready handoffs. Start a new idea with Design Coach, or diagnose an existing product with Hyper-Review.

Claude Code · PurplePrint MCP
Connected
I want to build an app that helps people living alone waste less food.
PurplePrint

Let’s hold the feature list for a moment. First, I’ll narrow down whose loss we are reducing and why this product should exist.

Next question How do they solve this today?

Product behavior example based on the live coaching contract.

Works now in Claude CodeCodexCursorVS CodeAntigravity

PURPLEPRINT SERVICE DESIGN COACH

One idea becomes a complete, build-ready blueprint.

Open one part of the system at a time. The detail stays available without turning the page into a wall of text.

purpleprint / design coach

PHASE 0–9

Service design comes before the feature list.

PurplePrint keeps every product decision in sequence, from the user and market problem to validation and build handoff. Each phase closes with an approved Markdown artifact.
WHAT REMAINSA complete, build-ready product blueprint

$ /purpleprint:start

 

▸ PHASE 0–9

 

Phase 0 Context why now and which market problem

Phase 1 Persona real users and current behavior

Phase 2 Problem journey · pains · competitors · value chain

Phase 3 Product core flow · MVP · revenue model

Phase 4 UX & IA principles · sitemap · journeys

Phase 5 UI design system · screen states

Phase 6 Review 39-axis recursive gate

Phase 7 GTM WTU · WTP · PMF experiments

Phase 8 Dev design stack · data · tracking · security

Phase 9 Handoff SSOT · build order · acceptance

 

→ A complete, build-ready product blueprint

HOW PURPLEPRINT DIFFERS

General AI agents force the choice back onto you. PurplePrint proposes three candidates first, then recommends the top choice with three grounds.

Claude (Sonnet 5)

PurplePrint

PURPLEPRINT INTELLIGENCES

Hyper Series

Design Coach establishes the product blueprint. Hyper engines investigate, challenge, and simulate that blueprint when deeper evidence or a stronger release decision is needed.

39 AXES · RECURSIVE REVIEW

The line between a plausible design and a build-ready one.

Before code, it grades input maturity and recursively checks 39 design axes. Missing screens, undefined tables, and broken flows return in severity order with a correction route.
  • 10-element maturity check
  • 39-axis recursive review
  • Severity-ranked correction

Actual PurplePrint output

purpleprint — feature preview

BUILDER GROWTH

The product improves. So does the builder.

my-growth shows how your agent direction changes across sessions: from accepting an answer to setting criteria, naming constraints, and asking for proof. It measures behavior patterns without sending conversation content to the server.

  • Session-by-session leadership trend
  • Nine-stage builder level
  • Content-free behavior records
View actual growth report

Actual my-growth output

purpleprint — my-growth

$ /purpleprint:my-growth

 

▸ Lead stage by session (content-free)

06-05 4·Orchestrator 0.35 directs dual-agent workflow

06-25 4~5 boundary 0.55 designs the strategy themselves

07-02 5·Human agent 0.62 articulates domain·journey

07-03 4~5 boundary 0.55 challenges real-vs-mock

 

trend following → directing·self-evolving

now 4~5 boundary (Orchestrator → Human agent)

next refute·redefine the AI's direction more → peer

 

▸ Builder level 6 → 7 / 9 · full report /en/growth

 

USER STORIES

The moments their product direction changed.

“Now I know what our team was missing.”

We'd already built the product and spent a million won on paid ads. Three of us were all in. But of 17 signups, the number who experienced the core feature was zero. We didn't know what was wrong, or how to fix it.

At this rate, we'd keep pouring in time and money for the same result. We even started wondering whether to keep going with the idea at all.

We ran Hyper-Review, and it found 45 gaps in our planning doc. Together with a founder consult on the report, the biggest problem turned out to be this: we'd built the product on a thin service design. The persona was off, there was no real user value-chain analysis — so the UX was a mess too. It diagnosed all of it in one hour, and corrected the exact next actions.

This kind of consulting normally takes weeks and millions of won. I had no idea you could get diagnosed and handed a fix this fast. In the end we saved months of time and tens of millions — and could finally build a real product on a real design. In just one hour. I love it.

Hyper-Review — B2C2B service · founder, ex-senior SI engineer · 3-person startup team

FAQ

What makes PurplePrint different?

How is this different from using Claude Code or Codex alone?

Claude Code and Codex are agents that execute work. PurplePrint runs inside them and adds the order of service-design decisions, evidence-backed recommendations, approval gates, artifacts, and the next action.

What is the biggest difference from a general AI agent?

PurplePrint does not stop at answering or hand the decision back to you. It proposes a leading option with alternatives and evidence, then keeps the next phase closed until the required design decisions are complete.

Can I start with only a one-line idea?

Yes. Design Coach begins by clarifying why the product should exist, then guides you through the user, problem, product, experience, go-to-market, technical design, and handoff in order.

Can I diagnose and improve a product I already built?

Yes. Start with Hyper-Review to find gaps in the current design, then return to the necessary Design Coach phases to strengthen them.

What do I actually get at the end?

Each approved phase leaves Markdown artifacts in your workspace, including context, personas, problem and product decisions, UX·IA·UI, GTM, technical architecture, data model, and a development handoff.

Where do my ideas, conversations, and design files live?

They stay in your workspace by default. The server remembers only content-free progress and usage state, not the content of your work.

Do I need another app or service-design expertise?

No. Keep working inside the agent you already use. You do not need to understand MCP, skills, hooks, memory, or compaction to follow the process.

START IN MINUTES

Install it with a prompt.

It registers the remote MCP, opens OAuth, and checks the connection for you.

Installation prompt for Claude Code · hover to preview Add the PurplePrint MCP to Claude Code. Server name: purpleprint, transport: http, URL: https://purpleprint-mcp.purpleprintai.workers.dev/mcp. Register it globally (user scope), and add alwaysLoad: true to mcpServers.purpleprint in ~/.claude.json (this prevents /purpleprint:* slash commands from going missing in recent Claude Code IDE). If editing the file directly, save as UTF-8 no BOM. Walk me through Authenticate in /mcp, and when done confirm that /purpleprint:start shows up. ⚠️ **Even if the /purpleprint:* slash commands never show up, it still works. Once connected, just say "start PurplePrint" in plain language — it runs the pp_start tool.** The slash list is a convenience, not the entry point. Don't reinstall or get stuck on it (Ctrl+Shift+P → "Developer: Reload Window" or an editor restart sometimes brings it back, but it's fine either way). If auth is needed and no prompt appears, disable and re-enable purpleprint, then check again. If it already shows "connected", auth is usually done — just say "start PurplePrint" to confirm.
Installation prompt for Cursor · VS Code · hover to preview Add the PurplePrint MCP to this editor. Server name: purpleprint, URL: https://purpleprint-mcp.purpleprintai.workers.dev/mcp. Register it as a remote/http MCP in this tool's settings — globally (user settings, e.g. ~/.cursor/mcp.json or VS Code user settings), not just one project. If an OAuth/authorize flow is needed, walk me through the browser login. ⚠️ **Even if the command list never shows up, it still works. Once connected, just say "start PurplePrint" in plain language — it runs the pp_start tool.** The list is a convenience, not the entry point — don't get stuck on it (reload window, restart, or a new thread sometimes brings it back, but it's fine either way). If auth is needed and no prompt appears, disable and re-enable purpleprint. If it already shows "connected", auth is usually done.
Installation prompt for Codex · hover to preview Add the PurplePrint MCP to Codex. Server name: purpleprint, URL: https://purpleprint-mcp.purpleprintai.workers.dev/mcp. Register it as a streamable HTTP MCP in ~/.codex/config.toml (shared by Codex CLI/IDE) and complete the OAuth login. On Windows PowerShell, if codex.ps1 is blocked by the execution policy, run it as cmd /c codex mcp ... When done, verify with codex mcp list. It's fine if /purpleprint:start doesn't appear in the / menu — just confirm that saying "start PurplePrint" lets you use the pp_start tool.
Installation prompt for Antigravity · hover to preview Add the PurplePrint MCP to Antigravity as a global MCP. Server name: purpleprint, URL: https://purpleprint-mcp.purpleprintai.workers.dev/mcp. Register it via serverUrl in ~/.gemini/config/mcp_config.json, and check ~/.gemini/antigravity/mcp_config.json if needed. After registering, go to Settings > Customizations > Installed MCP Servers, click Refresh, and complete Authenticate. When done, confirm that saying "start PurplePrint" invokes the MCP tool.

To start: after PurplePrint is connected, enter /purpleprint:start or say “Let's start PurplePrint.”

BETA

$13 / month

14 days free, no card. Design Coach, Hyper-Review, and builder growth tracking included.