managing-claude-code-meta

The managing-claude-code-meta skill is designed to facilitate the setup, management, and auditing of CLAUDE.md files within projects that adopt the promode methodology. It addresses the challenge of standardizing AI agent behavior documentation while maintaining a clear separation between main agent instructions and project-specific technical details. By ensuring that CLAUDE.md files are installed correctly, migrated properly, and audited for compliance, this skill helps maintain consistency, reduces errors, and enforces best practices in agent-driven software development workflows.

The skill provides several key features and capabilities. It supports the installation of standard CLAUDE.md files into new projects, ensuring that the main agent's behavior is defined consistently. For existing projects, it enables migration of non-promode CLAUDE.md content into AGENT_ORIENTATION.md, preserving project-specific details while adhering to the promode methodology. It also includes auditing capabilities to verify that projects follow progressive disclosure principles and maintain a single source of truth. Additionally, the skill enforces rules such as never modifying the standard CLAUDE.md, avoiding duplication between CLAUDE.md and AGENT_ORIENTATION.md, and escalating when conflicts or unusual conditions arise.

Typical use cases include setting up new AI agent-driven projects, refactoring existing CLAUDE.md files, auditing project configurations for compliance, and updating installations to the latest promode standards. The target users are developers, AI engineers, and project maintainers who work with multi-agent workflows and require structured, consistent guidance for main and sub-agent behavior. By using this skill, teams can ensure that their projects are aligned with best practices, maintain test-driven documentation, and facilitate smooth delegation and orchestration of AI agent tasks.