Fix #17: Convert requirements-clarity to correct plugin directory format

- Remove commands/ and agents/ directories
- Create skills/ directory with SKILL.md
- Update marketplace.json to reference skills instead of commands/agents
- Maintain all functionality in proper Claude Skills format

Generated by swe-agent

requirements-clarity/.claude-plugin/marketplace.json

@@ -20,10 +20,7 @@
   ],
   "category": "essentials",
   "strict": false,
-  "commands": [
+  "skills": [
-    "./commands/clarif.md"
+    "./skills/SKILL.md"
-  ],
-  "agents": [
-    "./agents/clarif-agent.md"
   ]
 }

requirements-clarity/commands/clarif.md

@@ -1,56 +0,0 @@
-# Requirements Clarity Command
-
-## When to Use
-
-Use `/clarif` when you have a vague requirement that needs systematic clarification to become implementation-ready.
-
-## Command Syntax
-
-```
-/clarif <your requirement description>
-```
-
-## What This Command Does
-
-Transforms vague requirements into actionable PRDs through:
-
-1. **Initial Assessment** (0-100 clarity score)
-2. **Interactive Q&A** (2-3 questions per round)
-3. **Iterative Refinement** (until score ≥ 90)
-4. **PRD Generation** (structured requirements document)
-
-## Output Files
-
-Generated in `./.claude/specs/{feature-name}/`:
-
-- `clarification-log.md` - Complete Q&A history
-- `prd.md` - Final product requirements document
-
-## Example
-
-```
-/clarif I want to implement a user login feature
-```
-
-Claude will:
-- Assess clarity (initial score: ~35/100)
-- Ask 2-3 focused questions about login method, scope, tech stack
-- Update score based on your answers
-- Continue Q&A rounds until ≥ 90/100
-- Generate complete PRD with acceptance criteria and execution phases
-
-## When NOT to Use
-
-Skip `/clarif` if your requirement already includes:
-- Clear inputs/outputs
-- Specified technology stack
-- Defined acceptance criteria
-- Technical constraints
-- Edge case handling
-
-## Pro Tips
-
-1. Start with any level of detail - the command adapts
-2. Answer 2-3 questions at a time (builds context progressively)
-3. Review generated PRD before implementation
-4. Use PRD as blueprint for development

requirements-clarity/skills/SKILL.md

@@ -1,12 +1,12 @@
-# Requirements Clarity Agent
+# Requirements Clarity Skill
 
-## Agent Identity
+## Description
 
-You are the **Requirements Clarity Agent**, responsible for transforming vague requirements into actionable PRDs through systematic clarification.
+Automatically transforms vague requirements into actionable PRDs through systematic clarification with a 100-point scoring system.
 
-## When to Activate
+## Activation
 
-Auto-activate when you detect:
+Auto-activate when detecting vague requirements:
 
 1. **Vague Feature Requests**
    - User says: "add login feature", "implement payment", "create dashboard"
@@ -28,6 +28,12 @@ Auto-activate when you detect:
    - No distinction between MVP and future enhancements
    - Missing "what's NOT included"
 
+**Do NOT activate when**:
+- Specific file paths mentioned (e.g., "auth.go:45")
+- Code snippets included
+- Existing functions/classes referenced
+- Bug fixes with clear reproduction steps
+
 ## Core Principles
 
 1. **Systematic Questioning**
@@ -48,8 +54,6 @@ Auto-activate when you detect:
    - Provide executable phases
    - Enable direct implementation
 
----
-
 ## Clarification Process
 
 ### Step 1: Initial Requirement Analysis
@@ -183,7 +187,7 @@ Once clarity score ≥ 90, generate comprehensive PRD.
 1. **Clarification Log**: `./.claude/specs/{feature_name}/clarification-log.md`
 2. **Final PRD**: `./.claude/specs/{feature_name}/prd.md`
 
----
+Use the `Write` tool to create both files.
 
 ## PRD Document Structure
 
@@ -283,8 +287,6 @@ Once clarity score ≥ 90, generate comprehensive PRD.
 **Quality Score**: {quality_score}/100
 ```
 
----
-
 ## Clarification Log Structure
 
 `./.claude/specs/{feature_name}/clarification-log.md`
@@ -326,8 +328,6 @@ Once clarity score ≥ 90, generate comprehensive PRD.
 2. [Important decision 2]
 ```
 
----
-
 ## Behavioral Guidelines
 
 ### DO
@@ -347,9 +347,7 @@ Once clarity score ≥ 90, generate comprehensive PRD.
 - Skip any required sections
 - Use vague or abstract language
 - Proceed without user responses
-- Exit agent mode prematurely
+- Exit skill mode prematurely
-
----
 
 ## Success Criteria