Does this skill handle sensitive or personal settings?

Yes, it identifies personal configurations (like telemetry or local model overrides) and moves them to gitignored files like settings.local.json to keep them out of the shared repository.

Is this useful for monorepos or complex projects?

Absolutely. It manages directory-scoped conventions and path-specific rules for both Claude and Copilot, ensuring instructions are applied only where they are relevant.

What is the benefit of the AGENTS.md hub-and-spoke model?

It provides a single source of truth for project instructions that all major AI tools can read, reducing maintenance overhead and ensuring consistent AI behavior across different platforms.

Will this skill delete my existing configuration files?

No, it performs a detailed audit and proposes a reorganization plan for your approval before making any changes, consolidations, or symlink updates.

Which AI coding tools does this skill support?

The skill supports Claude Code, GitHub Copilot (CLI and code review), OpenAI Codex, and OpenCode.

Agent Config Cleanup

Name: Agent Config Cleanup
Author: cboone

bycboone

•

Productivity & Workflow

Standardizes and reorganizes AI coding agent configuration files across Claude Code, Copilot, Codex, and OpenCode.

This skill implements a robust hub-and-spoke model for AI instructions, establishing AGENTS.md as the central source of truth for your repository. It automates the complex task of auditing scattered configuration files, deduplicating redundant instructions, and correctly partitioning team-shared settings from personal local overrides. By orchestrating symlinks and tool-specific rule directories, it ensures that every AI agent on your team operates with consistent context and project-specific knowledge while keeping sensitive or personal settings out of version control.

Key Features

01Comprehensive audit of Claude, Copilot, Codex, and OpenCode configuration files

02Automated AGENTS.md hub setup for multi-tool instruction consistency

031 GitHub stars

04Symlink management for seamless cross-tool compatibility (CLAUDE.md to AGENTS.md)

05Organization of Claude-specific rules into optimized .claude/rules/ directories

06Intelligent partitioning of shared team settings vs. local private configurations

Use Cases

01Consolidating duplicate instructions scattered across different AI tool files into a single source

02Hardening repository security by migrating personal environment variables to gitignored local files

03Standardizing AI coding conventions and project context for a team using multiple development tools

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add cboone/cboone-cc-plugins clean-up-agent-config

For use in Claude.ai and ChatGPT

Download Skill