mirror of
https://github.com/catlog22/Claude-Code-Workflow.git
synced 2026-02-10 02:24:35 +08:00
## Major Changes ### Parameter Flexibility - Changed `<skill_name>` from Required to Optional `[skill_name]` - Updated argument-hint: `[skill_name] "task intent description"` - Description: "Activate SKILL package (auto-detect or manual)" ### Auto-Detection Strategy (New Feature) Added Step 1: Determine SKILL Name (if not provided) 1. **Path Extraction**: Scan task for file paths, extract project names - Example: `"D:\projects\my_project\src\auth.py"` → extracts `my_project` 2. **Keyword Matching**: Match task keywords against SKILL descriptions 3. **Validation**: Check if extracted name exists in `.claude/skills/` ### Core Philosophy Update - Changed from "Manual Activation" → "Flexible Activation" - Auto-detect from task description/paths OR user explicitly specifies ### Enhanced Documentation - Parameters section: Added auto-detection explanation and path example - Task intent description: Now serves dual purpose (auto-detection + scope selection) - Updated all execution flow steps (Step 1 → Step 2 → Step 3) ### Usage Examples - **Example 1**: Manual specification (explicit skill_name) - **Example 2**: Auto-detection from path (NEW) - Shows complete auto-detection workflow - Path extraction → Validation → Activation ## Benefits - ✅ **Smarter activation**: System extracts skill name from context - ✅ **Reduced user effort**: No need to specify obvious skill names - ✅ **Backward compatible**: Manual specification still supported - ✅ **Path-aware**: Intelligent extraction from file paths in task description 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude <noreply@anthropic.com>