AI-First Engineering Standards 🏎️🛡️

Governing the speed of AI with the precision of Engineering.

In the era of AI-First development, speed is no longer the bottleneck—governance is. While AI models can generate hundreds of lines of code per minute, without the proper guardrails, technical debt and security risks can accumulate at the same velocity.

"In Formula 1, going faster without appropriate guardrails is a recipe for a catastrophic crash. This repository aims to be the steering wheel that guides the AI engine."

Core Principles:

• Governance by Design: Don't just ask the AI to "be good." Enforce your architectural patterns.
• Tool Agnostic: Define rules once in src_rules/ and deploy them to Cursor, GitHub Copilot, or Claude.
• Scale: Ensure your AI assistants follow modern standards automatically.

---

🏗️ Repository Structure

• src_rules/: Contains pure Markdown files with specific engineering standards.
• templates/: Configuration templates for different AI tools (.mdc for Cursor, .txt for Claude, etc.).
• generate_rules.py: This Python script merges your rules with templates to generate ready-to-use config files.
• output/: The final artifacts, organized by tool, ready to be copied into your project.

---

Current Guardrails included:

Git Discipline — Verification steps & Conventional Commits.

Angular Modern Excellence — Signals, Resource API, Standalone.

React & Next.js — Hydration safety (no browser APIs during render), stable hook references with useCallback, monorepo tsconfig isolation.

AWS Serverless (Lambda, DynamoDB, SAM) — CORS headers on every response path, DynamoDB TTL as Unix epoch Number, correct SAM template structure for TTL.

Security — Never include credentials, secrets, or sensitive data in code, configuration files, or commit history. AI tools operate on what you give them — review every diff before committing, especially when working under time pressure.

---

🚀 How to Use & Collaborate

1. Fork & Customize

Every project and team is unique. I encourage you to fork this repository.

• Add your own project-specific rules to src_rules/.
• Customize the templates to fit your team's stack.
• Maintain your own fork as your team's internal governance center.

2. Standardize

Run the generator to stay up-to-date.

Generate all rules:

python3 generate_rules.py

Generate a specific rule only:

python3 generate_rules.py react-next

Note: Use python3 on Linux/macOS. On Windows, use python or py depending on your installation.

The generator produces ready-to-use artifacts for three AI tools:

Tool	Output path	How to use
Cursor	output/.cursor/rules/*.mdc	Copy .cursor/ to your project root
Claude Code	output/CLAUDE.md	Copy CLAUDE.md to your project root
GitHub Copilot	output/.github/copilot-instructions.md	Copy .github/ to your project root

3. Collaborate

PRs with rules that proved useful in your projects are welcome 🫶