matura - Agentic AI SDLC Planner

matura transforms a rough software idea into a complete, implementation-ready software plan using Agentic AI. Built for the Swiss Life Claude Builders Hackathon.

The system guides users through the full SDLC workflow:

Idea → Vision → Requirements → Architecture → Backlog → Tests

All artifacts stay aligned through a retro-feedback loop — changes at any stage propagate across all others automatically.

Tech Stack

• Frontend: Next.js 16 (App Router), React 19, Tailwind CSS v4
• Backend: Convex (real-time database + server functions)
• AI Engine: Claude (Anthropic SDK) — Sonnet 4, Sonnet 4.6, Haiku 4.5
• Auth: Clerk (keyless mode)
• Visualization: React Flow (architecture diagrams)

Prerequisites

• Node.js 20+
• pnpm (npm install -g pnpm)
• An Anthropic API key
• A Convex account (free at convex.dev)

Setup & Run

# 1. Clone the repository
git clone <repo-url>
cd anthropic_hackathon

# 2. Install dependencies
pnpm install

# 3. Configure environment variables
cp .env.template .env.local
# Then add your Anthropic API key in .env.local:
#   ANTHROPIC_API_KEY=sk-ant-...
# All other variables (Convex, Clerk) are pre-configured in the template.

# 4. Start Convex backend (in one terminal)
npx convex dev

# 5. Start Next.js dev server (in another terminal)
pnpm dev

# 6. Open http://localhost:3000

Project Structure

app/
  page.tsx              # Landing page with Clerk sign-up
  home/page.tsx         # Dashboard — list all project flows
  onboarding/page.tsx   # Guided onboarding wizard (Stage 1)
  workspace/page.tsx    # Artifact generation & refinement (Stage 2 + 3)
  api/
    generate/           # SDLC artifact generation pipeline
    feedback/           # Retro-feedback alignment loop
    generate-architecture/  # Agentic architecture generation (tool use)
    onboarding-prefill/ # AI-powered onboarding field suggestions
    field-refine/       # Inline field refinement via streaming

convex/
  schema.ts             # Database schema (projects, artifacts, onboarding, etc.)
  projects.ts           # Project CRUD operations
  artifacts.ts          # Artifact storage and retrieval
  onboarding.ts         # Onboarding data persistence
  documents.ts          # Internal knowledge base (for architecture agent)
  techSpecs.ts          # Legacy tech specs generation
  seed.ts               # Knowledge base seed data

components/
  ArchGraph.tsx         # Interactive React Flow architecture diagram
  ProsePanel.tsx        # Streaming prose explanation panel
  TechStackLegend.tsx   # Architecture diagram legend
  MaturaLogo.tsx        # Brand logo component

Available Scripts

pnpm dev          # Start Next.js dev server (localhost:3000)
pnpm build        # Build for production
pnpm start        # Start production server
pnpm lint         # Run ESLint
npx convex dev    # Start Convex dev server (run in parallel)
npx convex deploy # Deploy Convex functions to production

How It Works

1. Sign up / Sign in at the landing page
2. Create a new flow from the home dashboard
3. Onboarding — describe your software idea; the AI pre-fills onboarding fields (user roles, workflows, approvals, notifications)
4. Workspace — the system generates all SDLC artifacts (Vision, Requirements, Architecture, Frameworks, Backlog, Tests, Cost Estimate, Competitive Analysis) using Claude
5. Refine — click any artifact card, add a refinement instruction, and the feedback loop updates all affected artifacts automatically
6. Architecture view — an agentic tool-use flow queries the internal knowledge base and produces an interactive architecture diagram

Team

Built by Alkinoos, Jannis, and Damiano for the Swiss Life Claude Builders Hackathon 2025.