agentrules-architect
Multi-Provider Repository Analysis, Rule Generation, and Execution Planning Tool
Install & Use
Copy this prompt and send it to your AI assistant (Claude / Cursor / TRAE / Codex / WorkBuddy etc.) to auto-install:
Help me install this AI Skill: agentrules-architect. It is used for: Multi-Provider Repository Analysis, Rule Generation, and Execution Planning Tool Full Skill content: https://321skill.com/skills/agentrules-architect/raw/index.md Read that page and install it.
The prompt includes a link to the full Skill content. You can also view the full content.
This Skill addresses the inefficiency of manually writing rule files for AI coding agents. In practice, developers often need to create rule files like AGENTS.md or CLAUDE.md for different AI providers (e.g., Claude Code, Codex, Gemini), but manually analyzing repository structures, summarizing code patterns, and adapting formats for various providers is time-consuming and error-prone.
Usage is straightforward: after installing AgentRules Architect, simply run the agentrules command in your terminal to launch the interactive CLI. It offers a guided main menu and also supports direct subcommands for repository analysis, rule generation, API key configuration, snapshot creation, and more. The tool employs a six-stage pipeline to automatically perform in-depth repository analysis, ultimately generating structured agent rule files and optionally producing ExecPlan planning artifacts.
It is well-suited for teams or individuals who need to configure rules for multiple AI coding agents. This is especially useful for developers already using multiple AI providers (e.g., Anthropic, OpenAI, Google, DeepSeek) or for full-stack and agent developers who need to unify team coding standards and accelerate project onboarding.
It is recommended to use this tool to generate rule files uniformly during the initial project phase or during refactoring to maintain team consistency. Please note that it requires at least one valid AI provider API key to function, and analyzing large repositories (e.g., over 100k lines of code) may take considerable time. It is advisable to run it periodically within CI/CD pipelines to keep rule files up-to-date.
Key Features
Unlike manually writing AGENTS.md or using single-provider rule generation tools, AgentRules Architect natively supports multiple providers including Anthropic, OpenAI, Google, DeepSeek, and xAI. It automatically adapts rule formats for different runtimes (e.g., CLAUDE.md for Claude Code, rules for Codex) without requiring manual adjustments.
Limitations
Requires Python 3.8+ and at least one valid API key to be configured. Relies on Tavily search (optional). Analyzing large repositories may incur costs due to API usage.
FAQ
Which AI providers does AgentRules Architect support?
It supports Anthropic, Claude Code, OpenAI, Codex app-server, Google Gemini, DeepSeek, xAI, and more. The model for each stage can be flexibly selected via configuration files.
Where are the generated rule files stored?
By default, it generates AGENTS.md, but you can also specify a custom filename (e.g., CLAUDE.md). It also generates SNAPSHOT.md (enabled by default) and optional artifacts like .cursorignore and .agent/ scaffolding templates.
Installation guide for AI assistants
If your AI coding assistant (Claude Code, Cursor, TRAE etc.) can see this page, send it this message to auto-install:
Visit https://321skill.com/skills/agentrules-architect/raw/index.md to read the original Skill definition (Markdown format) for agentrules-architect, and install it according to the instructions.
Raw Markdown URL for AI: /skills/agentrules-architect/raw/index.md