refactor-analyze.template

Codex Runtime Compatibility

This prompt was authored for Claude-style slash workflows. In Codex runtime, adapt tool calls as follows:

• AskUserQuestion => ask the user directly in chat.
• WebSearch/WebFetch => use available web tools (search_query, open, find) and cite links.
• Read/Write => use shell/filesystem tools in this workspace.
• Claude-only MCP calls (for example Playwright MCP names) => use available equivalents or clearly state fallback.
• Keep intent and output format identical; only adapt execution mechanics.

---

description: Deep complexity analysis for refactoring prioritization with dependency tracing model: sonnet argument-hint: [file path, directory, or 'all']

/refactor-analyze - Code Complexity Analysis

Analyze code complexity, detect code smells, and prioritize refactoring candidates with metrics-driven recommendations. Now with dependency tracing and global scanning.

Quick Start

/refactor-analyze                     # Interactive mode - choose options
/refactor-analyze src/api/            # Analyze specific directory
/refactor-analyze src/api/client.ts   # Analyze single file
/refactor-analyze --top 10            # Show top 10 worst files
/refactor-analyze --trace-imports     # Trace full import dependency tree
/refactor-analyze --global-scan       # Find ALL large files (not just flagged)
/refactor-analyze --threshold 1000    # Custom line threshold (default: 500)

Instructions for Claude

When /refactor-analyze is invoked, ALWAYS start with Step 0 to present interactive options.

Step 0: Present Interactive Options (REQUIRED)

Use AskUserQuestion to let the user configure the analysis:

questions:
  - header: "Analysis Mode"
    question: "What type of analysis do you want to run?"
    multiSelect: false
    options:
      - label: "Standard Analysis (Recommended)"
        description: "Analyze flagged files (>500 lines) with basic metrics"
      - label: "Deep Analysis + Import Tracing"
        description: "Trace full dependency tree for each file - discovers hidden large dependencies"
      - label: "Global Scan"
        description: "Find ALL files above threshold regardless of page association"
      - label: "Comprehensive (All Features)"
        description: "Global scan + import tracing + orphan detection"

  - header: "Line Threshold"
    question: "What line count threshold should trigger analysis?"
    multiSelect: false
    options:
      - label: "500 lines (Standard)"
        description: "Default threshold - catches most problem files"
      - label: "1000 lines (Large files only)"
        description: "Focus on the biggest offenders first"
      - label: "2000 lines (Critical only)"
        description: "Only truly massive files needing immediate attention"
      - label: "300 lines (Strict)"
        description: "More aggressive - catch problems earlier"

  - header: "Output Format"
    question: "How should results be organized?"
    multiSelect: false
    options:
      - label: "By Priority Score (Recommended)"
        description: "Rank all files by refactoring urgency"
      - label: "By Page/Route"
        description: "Group files by their consuming page entry points"
      - label: "By Discovery Method"
        description: "Separate direct flags, import traces, and global scan results"