AI Agents and Rules (github.com)

An open-source project called Agents and Rules has launched (or is being announced) as a curated, community-driven collection of AI-readable coding standards, best practices, and rule sets for developers and cloud/platform engineers. Designed to act as a knowledge base for AI coding assistants like Kiro, Cursor, GitHub Copilot and others, the repo already includes Terraform Rules for clean, secure IaC and promises more languages and specialized agents soon. Contributors get recognition (name + GitHub profile), and the project encourages PRs, simple friendly rule files, icons (drop SVGs in rules/icons/), and practical examples. Technically the rules are authored in an AI-optimized format with frontmatter metadata (name, description, category, icon, version, globs), clear good-vs-bad examples, runnable code snippets, and organized sections per language/tool—making them suitable as steering files in AI IDEs, references for code reviews, or team coding standards. The approach matters because it standardizes how AI assistants reason about style and safety, improves maintainability across teams, and enables seamless integration into developer workflows while remaining extensible through community contributions.