Ki Startup Command Center

Development environment for Ki - the world's first Human-AI-Human relationship intelligence platform. A lean 3-person startup building breakthrough AI technology that processes both partners simultaneously while maintaining privacy.

Overview

This command center orchestrates a complex multi-repository architecture including:

**Ki Platform**: Next.js 15 Turborepo monorepo with LangGraph AI engine

**Business Operations**: Data room, planning, customer development, fundraising

**Automation**: n8n workflows for business processes

Built on **Next Forge** - a production-ready Next.js boilerplate with enterprise-grade tooling.

Repository Architecture

**Main Repository**: `ki-startup-command-center`

**Submodules**:

`ki-platform` - Main platform (Next.js monorepo + LangGraph backend)

`ki-business` - Business operations and investor materials

`ki-automation` - n8n business automation workflows

Core Commands

Essential Development Commands

Use the Makefile for all primary operations:

```bash

Complete setup (submodules + Ki platform + automation)

make setup

Start complete development environment

make dev

Setup Ki platform specifically

make ki-setup

Stop all services

make stop

Open n8n automation dashboard (admin/ki2024)

make n8n

Backup everything

make backup

Clean up everything

make clean

```

Git Automation System (⭐ Claude Code Compatible)

Ki includes comprehensive git automation that handles all submodule repositories automatically:

```bash

Check status of all submodules and main repository

make git-status

Push changes in all submodules + main repository

make git-push

Pull latest changes from all repositories

make git-pull

```

**Key Benefits**:

Single command operations across all repositories

Automated commit messages with timestamps and proper attribution

Intelligent status reporting showing detailed information

Error handling for missing submodules and connection issues

Fully compatible with Claude Code execution

**Direct Script Execution**:

```bash

./scripts/git-submodule-status.sh # Check all repository status

./scripts/git-submodule-push.sh # Push all changes

./scripts/git-submodule-pull.sh # Pull all updates

```

Service URLs (when running `make dev`)

**Ki App**: http://localhost:3001 - Main relationship intelligence platform

**AI Engine API**: http://localhost:8000 - FastAPI with LangGraph backend

**Marketing Site**: http://localhost:3000 - Next.js website

**n8n Automation**: http://localhost:5678 - Business automation workflows

Next Forge Monorepo Commands

Navigate to Ki platform for monorepo-specific commands:

```bash

cd submodules/product/ki-platform

```

**Core Development**:

```bash

pnpm dev # Start all Next.js apps in development mode

pnpm build # Build all applications for production

pnpm test # Run Vitest tests across all packages

pnpm lint # Ultracite linting (Biome-based, faster than ESLint)

pnpm format # Format code with Ultracite

pnpm typecheck # TypeScript checking across monorepo

pnpm analyze # Bundle analysis for optimization

pnpm translate # Internationalization workflow

pnpm boundaries # Enforce package boundary rules

```

**Database Operations**:

```bash

pnpm migrate # Prisma format + generate + push to database

```

**Package Management**:

```bash

pnpm install # Install dependencies (preferred package manager)

pnpm bump-deps # Update all dependencies

pnpm bump-ui # Update shadcn/ui components

pnpm clean # Clean node_modules across workspace

```

**Backend Testing**:

```bash

cd apps/langgraph-backend

pytest tests/

python -m pytest tests/ -v

```

MCP Server Configuration

This project includes comprehensive MCP (Model Context Protocol) server configuration in `.mcp.json`:

```bash

View all configured MCP servers

claude mcp list

```

**Available MCP Servers**:

**puppeteer**: Web automation for competitive analysis and testing

**filesystem**: Enhanced file operations for data room management

**github**: Repository management across ki-platform, ki-business, ki-automation

**postgres**: Database operations for user analytics and metrics

**gdrive**: Cloud storage integration for investor documents

**slack**: Team communication automation

**calendar**: Meeting scheduling and milestone tracking

**analytics**: Business intelligence and user behavior analysis

MCP servers provide automated capabilities for web automation, database operations, file management, team collaboration, and repository management.

Architecture Overview

Core Innovation: Human-AI-Human Framework

Ki's breakthrough architecture processes both partners simultaneously while maintaining privacy:

```

Partner A → Private Channel → Ki AI Core ← Private Channel ← Partner B

↓

Empathy AI Layer

↓

Personalized Dual Responses

```

LangGraph AI Engine

Core relationship intelligence workflow in `submodules/product/ki-platform/apps/langgraph-backend/src/graphs/ki_relationship_graph.py`:

**Processing Flow**:

```

intake → safety_check → emotional_analysis → conflict_detection

→ anxiety_detection → empathy_processing → pattern_recognition

→ response_generation → memory_update

```

**Key Components**:

Dual-partner message processing with separate private channels

Emotional intelligence with empathy AI layer

Conflict detection using Thomas-Kilmann styles

Anxiety detection and safety monitoring

Pattern recognition for relationship dynamics

Personalized responses for each partner

Technology Stack (Next Forge Architecture)

**Frontend & UI**:

Next.js 15 with App Router (React 19.1.0)

Radix UI primitives with shadcn/ui design system

Tailwind CSS 4.1.7 with class-variance-authority

React Hook Form + Zod validation

Geist font family, next-themes for dark/light mode

**Backend & Infrastructure**:

LangGraph + OpenAI GPT-4 + Anthropic Claude

FastAPI (Python 3.11+) + Next.js API routes

PostgreSQL with Prisma ORM + Neon serverless

Redis + WebSocket support

Clerk authentication

Integrated cloud storage

**Development & DevOps**:

Turborepo for build orchestration

pnpm 10.11.0 package manager

TypeScript 5.8.3 with strict configuration

Ultracite (Biome-based) + ESLint for linting

Vitest for testing

n8n workflows for business automation

Docker Compose orchestration

**Production & Monitoring**:

Vercel deployment with edge functions

PostHog + Vercel Analytics + Google Analytics

Sentry with performance monitoring

Logtail for structured logging

Resend with React Email templates

Stripe payments with agent toolkit

Repository Structure

```

submodules/

├── product/ki-platform/ # Main Ki platform (Next.js Turborepo monorepo)

│ ├── apps/web/ # Marketing website (Next.js 15)

│ ├── apps/app/ # Ki relationship app (Next.js 15)

│ ├── apps/docs/ # Documentation site (Next.js 15)

│ ├── apps/langgraph-backend/ # LangGraph AI engine (Python + FastAPI)

│ ├── packages/ui/ # Shared React component library

│ ├── packages/eslint-config/ # Shared ESLint configurations

│ └── packages/typescript-config/ # Shared TypeScript configurations

├── business/ # Complete business operations

│ ├── data-room/ # Investor documentation for $1.5M pre-seed

│ ├── planning/ # MVP roadmap and weekly goals

│ ├── operations/ # Team sync, metrics, automation

│ ├── growth/ # Customer development, market research

│ ├── customer-research/ # User interviews and validation

│ └── fundraising-prep/ # Investment materials and prep

└── automation/n8n-workflows/ # Business automation workflows

```

Development Guidelines

Next Forge Architecture Patterns

**Environment Variables (Type-Safe)**:

Each package defines environment variables in `keys.ts` files

Uses @t3-oss/env-nextjs for runtime validation and autocompletion

Server/client separation with Zod validation schemas

Update both `.env.local` files and relevant `keys.ts` when adding env vars

**Package Architecture**:

Each feature is a workspace package (e.g., `@repo/auth`, `@repo/database`)

Shared dependencies via common TypeScript configs and UI components

Use workspace aliases (`@repo/*`) for internal packages

Auth, database, email, payments are isolated packages

**UI Component Development**:

Use `@repo/design-system` with Radix UI primitives

Tailwind CSS with class-variance-authority for component variants

Built-in dark/light mode support with next-themes

Radix UI ensures WCAG compliance

Geist font family for consistent branding

**Data Layer Best Practices**:

Prisma with type-safe database operations

Zod schemas for runtime type checking

Use `pnpm migrate` for database schema updates

Neon serverless PostgreSQL with connection pooling

Privacy-First Development

All relationship data is highly sensitive and encrypted

Maintain strict separation between partners' private data channels

Follow HIPAA compliance patterns for healthcare-grade security

Never log or expose personal relationship content

Use `@repo/security` for consistent protection patterns

AI/ML Best Practices

**Emotional Intelligence**: UI adapts to detected emotional states

**Pattern Recognition**: Identify dynamics without clinical labeling

**Voice-First**: Seamless voice-visual interaction

**Privacy**: All AI processing respects partner privacy boundaries

Workflow Patterns

Standard Development Workflow

1. **Start development environment**:

```bash

make dev

```

2. **Work on specific applications**:

```bash

cd submodules/product/ki-platform

pnpm dev

```

Apps run on:

- Web (marketing): http://localhost:3000

- App (platform): http://localhost:3001

- Docs: http://localhost:3002

- Storybook: http://localhost:6006

3. **Add new features following Next Forge patterns**:

- Use workspace packages (`@repo/*`) for shared logic

- Follow environment variable patterns with `keys.ts`

- Implement UI with design-system components

- Add comprehensive tests with Vitest

4. **Quality assurance**:

```bash

pnpm lint # Fast linting with Ultracite

pnpm typecheck # Ensure type safety

pnpm test # Run test suite

pnpm analyze # Check bundle sizes

```

5. **Database changes**:

```bash

pnpm migrate # Handle schema updates safely

```

Git Workflow with Automation

1. **Check status across all repositories**:

```bash

make git-status

```

2. **Make changes in any submodule or main repository**

3. **Push all changes with single command**:

```bash

make git-push

```

This automatically:

- Commits changes in each submodule with timestamped messages

- Pushes submodule changes to their respective remotes

- Updates submodule references in main repository

- Pushes main repository changes

4. **Pull latest updates**:

```bash

make git-pull

```

Key Files and Locations

**LangGraph AI Engine**: `submodules/product/ki-platform/apps/langgraph-backend/src/graphs/ki_relationship_graph.py`

**Environment Config**: `config/` directory and `.env.local` files in each app

**Business Operations**: `submodules/business/` directory

**Data Room**: `submodules/business/data-room/` for investor documentation

**Automation Workflows**: `submodules/automation/n8n-workflows/`

**Git Automation Scripts**: `scripts/git-submodule-*.sh`

Important Notes

**Package Manager**: Always use `pnpm` for dependency management in the monorepo

**Git Automation**: Prefer `make git-*` commands over manual git operations

**Privacy**: Never log or expose relationship data; maintain strict channel separation

**MCP Servers**: Leverage configured MCP servers for enhanced capabilities

**Next Forge Patterns**: Follow established patterns for environment variables, package structure, and UI components

Ki Startup Command Center

Ki Startup Command Center

Overview

Repository Architecture

Core Commands

Essential Development Commands

Complete setup (submodules + Ki platform + automation)

Start complete development environment

Setup Ki platform specifically

Stop all services

Open n8n automation dashboard (admin/ki2024)

Backup everything

Clean up everything

Git Automation System (⭐ Claude Code Compatible)

Check status of all submodules and main repository

Push changes in all submodules + main repository

Pull latest changes from all repositories

Service URLs (when running `make dev`)

Next Forge Monorepo Commands

MCP Server Configuration

View all configured MCP servers

Architecture Overview

Core Innovation: Human-AI-Human Framework

LangGraph AI Engine

Technology Stack (Next Forge Architecture)

Repository Structure

Development Guidelines

Next Forge Architecture Patterns

Privacy-First Development

AI/ML Best Practices

Workflow Patterns

Standard Development Workflow

Git Workflow with Automation

Key Files and Locations

Important Notes

Reviews (0)