agent: update make-task-issue to use prompt-based input instead of editor

[p-iknow]

Context

The /make-task-issue command currently uses an editor-based input approach, which doesn't work properly in some environments or requires complex editor configuration. This degrades the user experience.

Codebase analysis findings:

• .claude/commands/make-task-issue.md:49-125 - Currently uses ${EDITOR:-vi} to create temporary file and open editor
• .claude/commands/make-command.md:47-100 - Already successfully uses prompt-based approach (Q1-Q4 interactive questions)
• Error handling section mentions "If editor cannot be opened, fall back to multi-line prompts" but not implemented

Requirement

Update the /make-task-issue command to use a prompt-based input approach, allowing users to input each section through interactive prompts.

Key requirements:

• Remove editor dependency
• Provide type-specific sequential questions
• Validate required sections (Context, Requirement)
• Allow optional sections to be skipped
• Maintain AI Enhancement functionality

Solution

Implement following /make-command's prompt-based pattern:

1. Section-by-section sequential input:

• Context (required): "Describe the background: Why is this needed? What is the current situation?"
• Requirement (required): "Describe what needs to be done and the expected outcome:"
• Solution (optional): "If you have ideas about how to solve this, describe the approach: (Enter to skip)"
• Test Plan/Verification/Content Outline (optional, type-specific): "How should this be tested or validated? (Enter to skip)"

2. Type-specific question mapping:

• feat, config, agent: Context → Requirement → Solution → Test Plan
• doc: Context → Requirement → Content Outline → Verification
• test: Context → Requirement → Test Scope → Verification
• perf, refactor: Context → Requirement → Solution → Verification
• fix: Context → Requirement → Solution → Test Plan

3. Implementation changes:

• Replace entire Section 2 "Collect User Input (Editor-based)"
• Remove Write tool logic for temporary file creation
• Remove Bash(${EDITOR:-vi}) calls
• Collect each section via interactive prompts
• Validate required fields (re-prompt if empty or whitespace only)

4. Maintain AI Enhancement flow:

• After user input collection completes
• Gather related context from codebase
• Present AI suggestions to user for approval

Test Plan

1. Test issue creation with various types:

   • /make-task-issue feat - Fill all sections
   • /make-task-issue doc - Fill required only, skip optional sections
   • /make-task-issue agent - Test AI enhancement approval/rejection

2. Verification items:

   • Confirm re-prompt when required sections are empty
   • Confirm optional sections can be skipped
   • Verify correct questions shown per type
   • Verify AI enhancement flow works properly
   • Verify final issue structure and content

Reference

• .claude/commands/make-task-issue.md - Current implementation
• .claude/commands/make-command.md - Reference prompt-based pattern