Scientia Superpowers

Full-stack AI engineer toolkit for the Scientia Capital ecosystem.

Features

Commands

• /deploy - Deploy to Vercel, Railway, or Docker
• /migrate - Run Supabase migrations with validation
• /switch-project - Quick context switch between 98+ projects
• /revenue-focus - Open tier-1 and tier-2 revenue projects
• /demo-prep - Prepare project demos
• /cost-check - Analyze LLM costs across projects
• /sync-env - Sync environment variables across platforms

Skills

• deployment-patterns - Vercel/Railway/Docker best practices
• supabase-mastery - Migrations, RLS, pgvector patterns
• cost-optimized-llm - NO OpenAI, tiered routing (DeepSeek → Haiku → Sonnet)
• revenue-acceleration - GTM workflows, demo prep, outreach

Agents

• deployment-validator - Proactively validates before deploy
• migration-checker - Validates SQL migrations for safety
• project-onboarder - Helps onboard to any project quickly

Hooks

• pre-deploy-check - Auto-validate before deployment
• claude-md-freshness - Remind when CLAUDE.md is stale

Installation

This plugin is installed globally at ~/.claude/plugins/scientia-superpowers.

Configuration

Create ~/.claude/scientia-superpowers.local.md to customize:

# Scientia Superpowers Settings

## Revenue Projects
- sales-agent
- dealer-scraper
- thetaroom
- ai-cost-optimizer
- signal-siphon
- fieldvault-ai
- gemma-payroll

## Default Deploy Platform
vercel

## Projects Root
/Users/tmkipper/Desktop/tk_projects

Tech Stack

• Frontend: Next.js 14+, React, TypeScript, Tailwind
• Backend: Python, FastAPI, Supabase
• AI: Claude, Gemini, DeepSeek (NO OpenAI)
• Voice: Twilio, Cartesia, Deepgram
• Deploy: Vercel, Railway, Docker

Critical Rules

1. NO OpenAI - Use Claude, Gemini, or DeepSeek only
2. API keys in .env only - Never hardcode
3. RLS everywhere - All Supabase tables must have RLS