Computer Lab

Output: Complete implementation with all stages documented.

Checkpoint Mode

"Run computer lab checkpoint mode"

I pause at each stage:

Stage 1: Tech Advisor
├─ [OUTPUT]
└─ "Approve to continue, or request changes"

Stage 2: UX Design
├─ [OUTPUT]
└─ "Approve to continue, or request changes"

... continues ...

You control the pace. Say "continue" to proceed, or give feedback to iterate.

Variation Mode

"Computer lab variations"

At key decision points, I generate options:

Stage 1: Tech Stack (3 variations)
├─ Option A: Next.js + Supabase + Stripe
├─ Option B: Astro + Firebase + Stripe
├─ Option C: Remix + PlanetScale + Stripe
└─ "Pick A, B, or C"

Stage 2: UI Approach (2 variations)
├─ Option A: Minimal, brutalist
├─ Option B: Rich, animated
└─ "Pick A or B"

You get choices. I build based on your selections.

The Full Sequence

┌─────────────────────────────────────────────────────────┐
│                      COMPUTER LAB                       │
├─────────────────────────────────────────────────────────┤
│                                                         │
│  1. DISCOVER                                            │
│     ├─ Tech Advisor → Research + Stack recommendation   │
│     └─ Research Librarian → Context gathering           │
│                                                         │
│  2. DESIGN                                              │
│     ├─ UX Librarian → User flows, IA                   │
│     ├─ Components Librarian → UI components            │
│     └─ Animation Librarian → Motion plan               │
│                                                         │
│  3. BUILD                                               │
│     ├─ Implementation Librarian → Execute the plan     │
│     ├─ Frontend Librarian → UI execution               │
│     ├─ Backend Librarian → API, data layer             │
│     └─ Database Librarian → Schema, queries            │
│                                                         │
│  4. QUALITY                                             │
│     ├─ Code Audit Librarian → Quality review           │
│     ├─ Testing Librarian → Tests                       │
│     └─ Performance Librarian → Optimization            │
│                                                         │
│  5. SECURE                                              │
│     └─ Security Librarian → Security audit             │
│                                                         │
│  6. SHIP                                                │
│     ├─ Exit Librarian → Final checklist                │
│     └─ Deployment → Go live                            │
│                                                         │
└─────────────────────────────────────────────────────────┘

How to Use

Starting The Lab

"Run computer lab. Build me a [describe your project]."

Options:
- "Run computer lab full auto" — No stops
- "Run computer lab checkpoint mode" — Pause at each stage
- "Computer lab variations" — Give me options at each stage
- "Computer lab checkpoint + variations" — Both

Example Prompts

"Run computer lab full auto. Build me a cannabis dispensary 
management SaaS with inventory tracking, compliance 
reporting, and payment processing."

"Computer lab variations. I need a landing page for 
my AI startup. Show me different design approaches."

"Run computer lab checkpoint mode. Build a dashboard for 
tracking creator analytics across TikTok and YouTube."

Stage Details

Stage 1: Discover

Librarians: Tech Advisor, Research

Actions:

• Lock context (what, who, constraints)
• Research current tech landscape
• Generate stack recommendation
• Pull relevant skills from library

Output: Stack decision + skill references

Stage 2: Design

Librarians: UX, Components, Animation

Actions:

• Define user flows
• Plan component architecture
• Decide animation approach
• Reference design philosophy

Output: Design plan + component list

Stage 3: Build

Librarians: Implementation, Frontend, Backend, Database

Actions:

• Execute implementation plan
• Build features per spec
• Set up data layer
• Connect integrations

Output: Working code

Stage 4: Quality

Librarians: Code Audit, Testing, Performance

Actions:

• Run quality checks (security, performance, code smells)
• Generate tests
• Optimize critical paths

Output: Audit report + fixes

Stage 5: Secure

Librarians: Security

Actions:

• Security audit
• Check auth flows
• Validate data handling
• Review env vars

Output: Security sign-off

Stage 6: Ship

Librarians: Exit, Deployment

Actions:

• Final checklist
• Pre-deploy verification
• Deploy

Output: Live application

Customizing The Sequence

You can skip or add stages:

"Run computer lab, but skip the security audit — 
this is just a prototype."

"Run computer lab and include the 3D librarian — 
I want WebGL elements."

"Run computer lab, design phase only — 
I'll build it myself."

Context Template

When activating Computer Lab, I gather:

## Lab Context

**Project:** [What you're building]
**Audience:** [Who it's for]
**Constraints:** [Timeline, budget, tech limitations]
**Mode:** [Full Auto / Checkpoint / Variation]
**Special Requirements:** [Compliance, integrations, etc.]

Related

• librarians/tech-advisor-librarian.md — Stack decisions
• librarians/facilitator-librarian.md — Library health
• ai-builder/agentic-workflows/SKILL.md — Under-the-hood patterns
• tech-stack/STACK-ROUTER.md — App type blueprints

Future: RAG Enhancement

The Lab can be enhanced with RAG retrieval:

1. Embed all skills in vector DB
2. User prompt → retrieve relevant skills automatically
3. Orchestrator chains based on retrieved context

See implementation_plan.md for roadmap.