cc-separation-guide

What problem does it solve?

Developers often face conflicts when using Claude Code alongside project-specific configurations or agent SDKs (like Google ADK, LangChain). This leads to confusion, incorrect skill activation, and ambiguous commands, hindering efficient AI assistance. This Skill provides a clear methodology to prevent these conflicts, ensuring seamless operation.

Core Features & Use Cases

• Namespace Isolation: Establishes clear prefixing conventions (cc- for meta, <project>- for domain) for skills, commands, and hooks, preventing naming collisions and improving clarity.
• Explicit Scope Declaration: Guides on using YAML frontmatter fields (scope, target) and CLAUDE.md hierarchy to define the context of each configuration, ensuring Claude understands the intent.
• Conflict Detection & Resolution: Explains how Claude detects ambiguities and provides strategies to disambiguate terms like "agent" or "hook" between tool and project contexts, ensuring the right action is taken.
• Use Case: A team is integrating Claude Code into a project that uses LangChain. This Skill helps them define lc- prefixes for LangChain-related skills and app- for project features, ensuring Claude doesn't confuse LangChain "tools" with Claude Code's built-in "tools," thus maintaining clear operational boundaries and avoiding misinterpretations.

Quick Start

Ask Claude: "How do I separate Claude Code configurations from my project's configurations?"