dev-rules-and-agents

Provides portable engineering rules and a 13-agent subagent roster for bootstrapping project AI configurations.

2|1|Updated Jun 23, 2026
One-click install
npx skills add https://github.com/Kaidanov/grekai-skills-4all --skill dev-rules-and-agents-kaidanov
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: dev-rules-and-agents
Source: https://github.com/Kaidanov/grekai-skills-4all/tree/main/skills/dev-rules-and-agents
Command: npx skills add https://github.com/Kaidanov/grekai-skills-4all --skill dev-rules-and-agents-kaidanov

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Teams repeatedly hand-write coding standards, refactoring methodologies, and subagent definitions for every new repository. This Skill packages proven, project-agnostic engineering rules plus a ready-made specialist agent roster so a new repo's CLAUDE.md, AGENTS.md, or .cursor rules can be bootstrapped from a known-good baseline instead of starting from scratch. ## Core Features & Use Cases - Portable rule sets: General coding rules, a behavior-preserving refactoring methodology, and session-hygiene/working-style rules that drop into any project's CLAUDE.md or AGENTS.md. - 13-agent roster: Specialist subagent definitions (frontend, backend, refactor, QA, Playwright, DevOps, security, UX, architecture, orchestration) in Claude-style frontmatter that down-converts to GitHub Copilot .agent.md files and Cursor prompts. - Routing standard: An architecture and standards guide defining activation order, agentic loops, token-usage standards, and a pruning standard for retiring agents. - Use Case: When starting a new Vite/React + .NET project, fold the rules files into the repo's CLAUDE.md, copy the agent definitions into .claude/agents/, and let the routing standard dispatch tasks like "add a rate limiter" to backend-dev, security-specialist, and qa-test-engineer automatically. ## Quick Start Ask your assistant to follow the rules in rules/coding-rules.md and rules/refactoring-rules.md, and to route tasks using the agent roster defined in agents/README.md.

Frequently Asked Questions about dev-rules-and-agents

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I add coding rules to CLAUDE.md for a new project?

Copy the relevant lines from rules/coding-rules.md, rules/refactoring-rules.md, and rules/working-style.md into your project's CLAUDE.md or AGENTS.md. The rules are project-agnostic, covering working style, code quality, safety, and documentation standards.

How do I create subagents for Claude Code and GitHub Copilot?

Copy the agent Markdown files into .claude/agents/ for Claude Code, or transform each file's frontmatter into .github/agents/<name>.agent.md for Copilot. Claude-style frontmatter is the richest format and down-converts cleanly to other tools.

Can I use these agent definitions with Cursor?

Cursor has no per-agent file format; it runs on .cursor/rules plus MCP. Reference the agent Markdown file in your prompt to get the same behavior, and fold the rules files into your .cursor/rules directory.

Are the 13 agents tied to a specific product?

The roster is an adaptable example extracted from a Vite/React + .NET XSLT data-mapping tool. Keep the structure and methodology, but swap product-specific nouns like paths, stack, and domain when reusing it on another project.

When should I use multiple agents versus one?

Default to one primary agent and add at most one specialist only after a real boundary is proven. The routing standard in agents/ARCHITECTURE_AND_STANDARDS.md defines activation order and discourages multi-agent bursts that waste tokens.