Security Rules for AI Coding Tools
Drop a rules file into your project and your AI coding tool enforces security patterns automatically. Works with Claude Code, Cursor, Windsurf, and GitHub Copilot.
5 stacks · 88 security rules · Free forever
1. Enter Your Email
We send a download link to your inbox. No account needed.
2. Drop In
Place the rules file at your project root. AI tools detect it automatically.
3. Code Safely
Your AI coding tool enforces security patterns on every generation.
Available Blueprints
Supabase + Stripe + Railway
Most PopularSecurity rules for full-stack SaaS applications built on Supabase (auth + database), Stripe (payments + subscriptions), and Railway (deployment). Covers the 9 most common vulnerabilities AI coding tools introduce.
Next.js + Prisma + Vercel
NewSecurity rules for full-stack applications built on Next.js (frontend + API routes), Prisma ORM (database access control), and Vercel (deployment). Covers SQL injection via raw queries, environment variable exposure, and unauthenticated server actions.
Next.js + Convex
NewSecurity rules for real-time applications built on Next.js (frontend) and Convex (backend database, server functions, file storage). Covers public function exposure, missing auth checks, data leaks in subscriptions, and schema validation.
Python + FastAPI + PostgreSQL
NewSecurity rules for backend APIs built on Python FastAPI with PostgreSQL. Covers SQL injection via f-strings and raw queries, authentication bypass, debug mode in production, hardcoded secrets, and insecure deserialization.
NemoClaw + OpenShell
NewSecurity rules for AI agent environments built on NVIDIA NemoClaw with OpenShell sandboxed execution. Covers policy files, network isolation, model routing, skill installation, credential management, and sandbox escape prevention.
More blueprints coming soon
Django + PostgreSQL + AWS · React + Firebase · SvelteKit + Supabase