One designer.
Full agency output.

Your entire design workflow, research to shipped code, running as AI skills you approve. Not prompts you write.

40 Skills 6 Workflow Agents 14 Knowledge Notes 6 MCP Integrations Running in 10 Minutes
Get Solo Design Studio — $99 USD See What's Inside
Solo Design Studio — Skill Pegboard
Design Cycle
Strategize 7
challenge-defaults
map-unknowns
reduce-principles
frame-vision
map-jobs
flip-argument
weigh-tradeoffs
Discover 6
audit-ux
scan-competitors
translate-analytics
extract-style
discover-assumptions
sketch-solutions
Refine 5
critique-design
review-design
audit-system
check-quality
plan-sprint
Ship 5
build-frontend
write-spec
define-motion
audit-accessibility
write-handoff
Reflect 4
write-casestudy
create-content
summarize-research
debrief-project
Support Systems
Leadership 5
prep-meeting
facilitate-session
draft-comms
frame-message
capture-minutes
Operations 4
run-standup
sync-notion
loop-skill
init-project
Knowledge 4
create-atoms
distill-notes
link-notes
route-insights
Claude AI Built for Claude Code · Works with any AI CLI
How it works

Running in 10 minutes.

Solo Design Studio — Interactive Setup ---------------------------------------
Scanning project...
✓ AI CLI detected (Claude Code)
✓ Obsidian vault found
✓ Figma MCP connected
✓ 40 skills loaded
✓ 6 agents configured
✓ Knowledge notes indexed
Ready. Type "what should I work on?" to begin.
You: What project moment am I in?
Studio: You have an active project "Checkout Redesign" with a spec draft from last week. Three paths:
1. critique-design → Review what you've built
2. write-spec → Finalize the spec
3. audit-ux → Audit the current live site
Recommended: critique-design
(Your spec has no review pass yet.)
Design Critique: Checkout Redesign ----------------------------------
Quality Gate Status
─────────────────────────
Info hierarchy ✓ Pass
Interaction model ✓ Pass
Edge cases ✓ Pass
Accessibility ⚠ 2 warnings
Responsive ✓ Pass
Score: 4.2 / 5.0
→ What's Next? Run write-spec to lock this down.
▶ Critiques
Critique — Checkout Redesign
Critique — Onboarding Flow
▶ Decisions
Single-page vs multi-step
Guest checkout default
▶ Knowledge
▶ Atoms
UX — Checkout Friction
Pattern — Progressive Disclosure
▶ Specs
Spec — Checkout Redesign
Critique — Checkout Redesign ×
Spec — Checkout Redesign ×
+
Critique — Checkout Redesign
tags critiquecheckoutQ1-2026
project Checkout Redesign
skill critique-design
score 4.2
Linked Decisions
  • Decision — Single-page vs multi-step
  • Decision — Guest checkout default
Knowledge Created
  • Atoms/UX — Checkout Friction Points
  • Atoms/Pattern — Progressive Disclosure
4 backlinks 2 new atoms Score: 4.2 / 5.0

I built this for myself. Then I realized every solo designer needs it.

Built over 1 year of daily use. Every skill tested on real projects with real deadlines. 40 skills · 6 agents · battle-tested
What's inside

40 skills. 5 phases. One cycle.

Every skill maps to a moment in your design project. Press a key. Get the output.

Strategize / Frame the problem before you design anything.
challenge-defaults
Kill assumptions before they become constraints
map-unknowns
Map what you know, don't know, and can't see
reduce-principles
Strip complexity to find what's irreducibly true
frame-vision
Align design work to business outcomes
map-jobs
Frame problems from user motivation instead of feature requests
flip-argument
Stress-test your position by making the opposite case
weigh-tradeoffs
Structured multi-criteria evaluation for hard choices
Discover / Understand the landscape.
audit-ux
Full UX teardown of any site. Competitive audit in 20 minutes
scan-competitors
Research and compare competitor design patterns
translate-analytics
Raw analytics data transformed into design implications
extract-style
Extract visual DNA from moodboards or live sites into design tokens
discover-assumptions
Assumption mapping + experiment design (Teresa Torres methodology)
sketch-solutions
Solution proposals from problem framing
Refine / Build with quality.
critique-design
Structured critique: info hierarchy, interaction efficiency, segmentation
review-design
Annotation-driven critique-fix-verify loop with pixel precision
audit-system
Design system consistency, naming, token usage. The hygiene pass you keep postponing
check-quality
Accessibility, interactions, performance, copywriting. The QA pass you skip
plan-sprint
Sprint planning, RICE/MoSCoW prioritization, capacity planning
Ship / Deliver artifacts.
build-frontend
Production-grade HTML/CSS/React from your specs. Design to code, no developer needed
write-spec
Design spec docs with component inventory, states, edge cases
define-motion
Web animation guidelines: easing, timing, performance, accessibility
audit-accessibility
WCAG compliance check and remediation plan
write-handoff
Component inventory, states matrix, edge cases, Definition of Done. Engineers stop asking questions
Reflect / Extract value between projects.
write-casestudy
Structure and write portfolio case studies from project work
create-content
Turn case studies into X threads, LinkedIn posts, newsletter sections
summarize-research
Progressive summarization. Research becomes a searchable library
debrief-project
Lessons learned, pattern extraction, what to carry forward

+ Leadership, Operations & Knowledge skills

prep-meeting
Pre-meeting intelligence: RACI mapping, power dynamics, framing script
facilitate-session
Session architecture, structured critique. Run the room
draft-comms
Design rationale, status updates, presentation scripts
frame-message
Executive communication for non-design stakeholders
capture-minutes
Meeting notes with action items, decisions, follow-ups
run-standup
Daily design standup: project state, today's focus, blockers
sync-notion
Push project artifacts to Notion for async collaboration
loop-skill
Run any skill in iterative passes for deeper analysis
init-project
Scaffold a new project: folders, briefs, stakeholder map
create-atoms
Turn insights into atomic knowledge notes
distill-notes
Progressive summarization of research
link-notes
Connect new knowledge to existing patterns
route-insights
Route wisdom from research to relevant skills
See it in action

Real output from real projects.

$ audit-ux cursor.com

UX Teardown: cursor.com
----------------------------------

Executive Summary
Cursor's site prioritizes developer trust
through live code demos over marketing copy.
3 critical usability gaps identified.

Information Architecture
| Page | JTBD | Score |
| Hero | Evaluate tool | 8/10 |
| Pricing | Compare plans | 6/10 |
| Docs | Learn to use | 7/10 |

Top 3 Recommendations
1. Add comparison table to pricing
2. Surface changelog in navigation
3. Reduce hero load time (4.2s)
Generated in 18 minutes
$ write-spec signup-flow

Design Spec: Signup Flow
----------------------------------

Component Inventory
| Component | States |
| Email input | 5 |
| Password field | 4 |
| OAuth buttons | 3 |
| Progress bar | 3 |

Edge Cases
- SSO redirect failure (timeout 30s)
- Duplicate email detection
- Password strength < minimum
- OAuth scope mismatch

Definition of Done
✓ All states documented
✓ Error copy finalized
✓ A11y requirements noted
Generated in 12 minutes
$ debrief-project checkout-redesign

Project Debrief: Checkout Redesign
----------------------------------
---
title: Debrief — Checkout Redesign
tags: [debrief, checkout, Q1-2026]
project: [[Checkout Redesign]]
status: complete
skills_used: [audit-ux, critique-design, write-spec]
---

Key Decisions
- [[Decision — Single-page vs multi-step]]
- [[Decision — Guest checkout default]]

Patterns to Carry Forward
1. Error prevention > error recovery
2. Progress indicators reduce abandonment 22%

Knowledge Created
- [[Atoms/UX — Checkout Friction Points]]
- [[Atoms/Pattern — Progressive Disclosure Forms]]

→ Linked to 4 existing notes. 2 new atoms created.
Generated in 8 minutes · Saved to vault
$ critique-design settings-page.fig

Design Critique: Settings Page
----------------------------------

Information Hierarchy
● Critical: Save button below fold on 768px
● Major: 6 toggles with no grouping
● Minor: Label contrast ratio 3.8:1

Interaction Efficiency
Clicks to change password: 7 (target: 4)
Clicks to update email: 5 (target: 3)

Recommendations
1. Sticky save bar at bottom
2. Group toggles by category
3. Increase label weight to 500
Generated in 6 minutes
$ prep-meeting design-review-q1

Meeting Prep: Q1 Design Review
----------------------------------

Context
3 projects shipping this quarter.
2 stakeholders need status updates.

Talking Points
1. Checkout redesign: 22% fewer drop-offs
2. Settings revamp: shipped 3 days early
3. Onboarding: blocked on API dependency

Risk Framing
● Onboarding: needs eng commitment by Feb 15
● Checkout: ready for A/B test next sprint

→ Draft agenda created. Linked to 3 project notes.
Generated in 4 minutes · Saved to vault
What this replaces

Stop paying for what you can own.

Lovart AI (design agent) $864/yr
Design subscription service $60,000/yr
Your billable hours lost to non-design work $52,000/yr
$99 USD

Launch price. Increases after first 100 buyers.
Once. Own it forever.

Get Solo Design Studio

Instant access after purchase.

Who it's for

Designers who ship.

Freelancer Full agency output. Research, design, code, docs. Less than 1 billable hour to own.
Senior IC designer 40% of your week is specs and audits. That's handled. Focus on craft.
Design team lead 10x your IC output. Run critiques, audits, sprint planning at scale.
Solo agency owner Compete with 5-person studios. Same output, one salary.
Design engineer Ship prototypes with real specs. Design and code in one flow.
Product designer From user research to shipped features. The full loop, automated.
Freelancer Full agency output. Research, design, code, docs. Less than 1 billable hour to own.
Senior IC designer 40% of your week is specs and audits. That's handled. Focus on craft.
Design team lead 10x your IC output. Run critiques, audits, sprint planning at scale.
Solo agency owner Compete with 5-person studios. Same output, one salary.
Design engineer Ship prototypes with real specs. Design and code in one flow.
Product designer From user research to shipped features. The full loop, automated.
How it compares

Not another AI tool.

DIY Prompting
You write every prompt from scratch
No memory between sessions
Generic output, no design methodology
Free (but hours of your time per task)
You are the system
No quality gates
Chat with AI
Generic AI Tool
One phase (usually generate)
Single role (design only)
Generate pixels faster
Subscription ($72+/mo forever)
Built by AI companies
No design methodology
Use a tool
Solo Design Studio
Full lifecycle: research to shipped code
8 roles with persistent project memory
Orchestrate the entire design process
One-time $99 USD. Yours forever
Built by a Product Designer who ships at scale
CCC framework with quality gates baked in
Hire an AI design lead
What you actually get

A complete operating system.

136 files. Running in 10 minutes.

Layer
What It Is
Why It Matters
40 Skills
Research to code to case study. Each encodes a senior-level process
The executable expertise
14 Knowledge Notes
Distilled design frameworks: critique methods, research methodology, token architecture, accessibility standards
Skills auto-load relevant knowledge. Design wisdom from production use
6 Agents
Workflow routing: Strategize, Discover, Refine, Ship, Reflect, Knowledge
Skills compose into end-to-end workflows
6 MCP Integrations
Figma, Playwright, Agentation, Excalidraw, Notion, Linear
Your tools, connected
Quality Gates
Hook scripts + skill-level verification on every output
Consistent quality on every run
FAQ

Questions you'd ask.

A prompt is a single instruction. Solo Design Studio is an operating system. 40 skills with structured protocols, quality gates, knowledge notes that auto-load design frameworks, a memory system that compounds across sessions, and progressive disclosure that chains skills into workflows. One is asking a question. The other is hiring a design lead who knows your projects, your stack, and your stakeholders.

No. It orchestrates the design process: research, strategy, specs, code, stakeholder comms, case studies. For visual design, it works with your Figma files through the Figma MCP integration: extracting styles, auditing systems, pushing annotations. You design. It handles everything around the design.

No. The build-frontend skill generates production-grade HTML/CSS/React, but you never need to write code yourself. Everything else is natural language in, structured output out. Research, specs, comms, critiques.

Depends on your CLI. For Claude Code: a Pro ($20/mo) or Max plan from Anthropic. For Codex: an OpenAI API key. For Gemini CLI: a Google AI key. The studio is markdown-based project instructions, so it runs on whichever provider your CLI connects to.

Built around Claude Code, but the system is structured markdown — CLAUDE.md, skills, agents, knowledge notes — so it works with any AI coding CLI that reads project instructions. Claude Code, Codex, Gemini CLI, Cursor, Windsurf. Larger models (Opus, GPT-4.1, Gemini 2.5 Pro) give the best results for complex skills like strategy and critique.

10 minutes. The interactive setup.sh script walks you through everything — MCP connections, project structure, memory system. Guided or self-directed, your choice.

No. Use what you have. Figma MCP and Playwright are the highest-value integrations. Figma for design system work, Playwright for live UX audits. Everything else is optional and can be added later.

The output works in any markdown editor. Obsidian-specific features (wikilinks, Dataview properties, Tasks plugin format) add power if you use Obsidian, but the content is standard markdown underneath. You won't miss anything critical without Obsidian.

Yes. Every skill has a TEMPLATE.md you can modify for output format, and a SKILL.md that defines the process. You can also create entirely new skills, add rules, and extend agents. The system is designed to grow with you.

The license covers one individual. Each team member needs their own copy. The outputs (specs, case studies, comms) are yours to share freely. The restriction is on redistributing the studio files themselves.

Yes. The core system — skills, agents, knowledge notes, persona — is structured markdown that any AI coding CLI can read. Codex, Gemini CLI, Cursor, Windsurf, OpenClaw. Hooks and MCP integrations (Figma, Playwright, Linear) are Claude Code-specific but the design workflow runs everywhere.

No. AI coding CLIs require an internet connection to reach their model APIs. The studio files themselves are local markdown on your machine. Your projects, memory, and knowledge never leave your device.

You get lifetime access to v1 and all v1.x updates (new skills, knowledge notes, bug fixes). Major versions (v2) will be offered at a significant upgrade discount.

A Product Designer who builds with AI daily. This is my actual workflow, 1 year of production use, packaged for you.

$99 USD

One-time. 40 skills. 6 agents. 14 knowledge notes. Yours forever.

Get Solo Design Studio — $99 USD
Tridip Thrizu
Tridip Thrizu
Product Designer · Building with AI daily
X LinkedIn