[FEATURE] /init command should instruct Claude to use relevant Skills before manual file exploration

Feature Request

Problem Statement

When the /init command is invoked to analyze a codebase and create a CLAUDE.md file, Claude currently defaults to manual file exploration (using Read, Glob, Bash tools) instead of leveraging available Skills that provide organized project knowledge.

This results in:

• Inefficient use of context by reading multiple files sequentially
• Missing synthesized knowledge that Skills already provide
• Longer analysis time to piece together architecture manually

Proposed Solution

Modify the /init command prompt to explicitly instruct Claude to:

1. Check for relevant Skills first before exploring files
2. Use Skills that match the analysis objective (e.g., project context gathering, architecture understanding)
3. Fall back to file exploration only for information not covered by Skills

Suggested addition to /init prompt:

Before exploring files manually, check if any available Skills provide organized 
project context or architectural knowledge. Use those Skills first to understand 
the codebase structure, then supplement with targeted file reads as needed.

Why This Matters

Skills align with Anthropic's documented use cases:

• "Execute company-specific data analysis workflows"
• "Automate personal workflows and customize Claude to match your work style"

The /init task (understanding architecture, identifying common commands, synthesizing documentation) directly matches what project-specific Skills are designed to provide.

Current Behavior

Claude receives the /init prompt asking to "analyze this codebase" and "understand high-level architecture," but doesn't recognize that available Skills (which synthesize this exact knowledge) should be the first tool to use.

Expected Behavior

Claude should:

1. Invoke relevant Skills to gather organized project knowledge
2. Use that synthesized information as the foundation for CLAUDE.md
3. Only explore individual files for specific gaps

Benefits

• More efficient context usage
• Faster CLAUDE.md generation
• Higher quality output leveraging pre-synthesized knowledge
• Better alignment with Skills' intended purpose