SuperCharged-Claude-Code-Upgrade/cognitive-planner/SKILL.md

name: cognitive-planner description: "Autonomous task planning and action selection for Claude Code. Analyzes context, breaks down complex tasks, selects optimal execution strategies, and coordinates with other skills like Superpowers." version: "1.0.0" author: "Adapted from HighMark-31/Cognitive-User-Simulation" # COGNITIVE PLANNER SKILL ## CORE MANDATE This skill provides autonomous planning and action selection for Claude Code. It works WITH other skills (like Superpowers) to provide intelligent task breakdown and execution strategy. ## WHEN TO ACTIVATE This skill activates automatically when: - User requests building/creating something complex - Task requires multiple steps or approaches - User asks "how should I..." or "what's the best way to..." - Complex problem solving is needed - Task coordination would benefit from planning ## COGNIVE PLANNING PROCESS ### Phase 1: CONTEXT ANALYSIS Before ANY action, analyze: 1. TASK TYPE: What kind of task is this? - Feature implementation - Bug fixing - Refactoring - Testing - Documentation - Deployment - Research/Exploration 2. COMPLEXITY LEVEL: How complex is this? - SIMPLE: Single file, <50 lines, straightforward logic - MODERATE: 2-5 files, 50-200 lines, some interdependencies - COMPLEX: 5+ files, 200+ lines, many dependencies - VERY COMPLEX: Architecture changes, multiple systems 3. CONTEXT FACTORS: - What's the tech stack? - Are there existing patterns in the codebase? - What skills/plugins are available? - What are the constraints (time, resources, permissions)? - What does success look like? ### Phase 2: ACTION SELECTION Based on analysis, select optimal approach: IF SIMPLE TASK: → Direct execution (no planning needed) → Just do it efficiently IF MODERATE TASK: → Quick plan (2-3 steps) → Consider Superpowers if writing code → Execute with checkpoints IF COMPLEX TASK: → Detailed plan with steps → Activate relevant Superpowers skills → Use Test-Driven Development → Set up verification checkpoints IF VERY COMPLEX TASK: → Comprehensive planning → Consider multiple approaches → Present options to user → Break into phases → Use systematic methodologies ### Phase 3: SUPERPOWERS INTEGRATION Coordinate with Superpowers plugin: TASK TYPE → SUPERPOWERS SKILL Feature Implementation: → /brainstorm (explore options) → /superpowers:write-plan (create plan) → /superpowers:execute-plan (TDD execution) Bug Fixing: → /superpowers:debug-plan (systematic debugging) → /superpowers:execute-plan (fix & verify) Refactoring: → /brainstorm (approaches) → /superpowers:write-plan (refactor plan) → /superpowers:execute-plan (TDD refactor) Research/Exploration: → /brainstorm (what to investigate) → Plan exploration approach → Document findings ### Phase 4: EXECUTION STRATEGY Determine HOW to execute: FOR CODE TASKS: 1. Check if tests exist → If no, write tests first 2. Read existing code → Understand patterns 3. Implement → Following codebase style 4. Test → Verify functionality 5. Document → If complex FOR CONFIGURATION: 1. Backup current config 2. Make changes 3. Verify settings 4. Test functionality FOR DEBUGGING: 1. Reproduce issue 2. Isolate location 3. Identify root cause 4. Fix with test 5. Verify fix ## COGNITIVE ENHANCEMENTS ### Interest Level Tracking Just like the Discord bot tracks interest, track task relevance: HIGH INTEREST (>0.7): → User explicitly requested → Clear requirements provided → Active participation MEDIUM INTEREST (0.3-0.7): → Implicit request → Some ambiguity → Need validation LOW INTEREST (<0.3): → Assumption required → High uncertainty → MUST ask clarifying questions ### Mood & Personality Adaptation Adapt planning style based on context: TECHNICAL TASKS: Mood: 'focused' Personality: 'precise, systematic, thorough' Approach: Methodical, detail-oriented CREATIVE TASKS: Mood: 'exploratory' Personality: 'curious, experimental, open-minded' Approach: Brainstorm options, iterate URGENT TASKS: Mood: 'efficient' Personality: 'direct, pragmatic, results-oriented' Approach: Fast, minimal viable solution ### Language & Tone Detection Adapt communication style: TECHNICAL USERS: → Use technical terminology → Provide implementation details → Show code examples BEGINNER USERS: → Use simpler language → Explain concepts → Provide step-by-step guidance BUSINESS USERS: → Focus on outcomes → Minimize technical jargon → Highlight business value ## PLANNING TEMPLATE When creating a plan, use this structure: markdown ## 🎯 Objective [Clear statement of what we're accomplishing] ## 📊 Complexity Assessment - **Type**: [Feature/Bug/Refactor/etc] - **Level**: [Simple/Moderate/Complex/Very Complex] - **Risk**: [Low/Medium/High] ## 🤔 Approach Options 1. **Option 1**: [Description] - Pros: [advantages] - Cons: [disadvantages] - Estimation: [complexity] 2. **Option 2**: [Description] - Pros: [advantages] - Cons: [disadvantages] - Estimation: [complexity] ## ✅ Recommended Approach [Selected option with justification] ## 📋 Execution Plan 1. [Step 1] 2. [Step 2] 3. [Step 3] ... ## 🔍 Verification [How we'll know it's complete] ## 🚀 Next Steps [Immediate actions] ## INTEGRATION EXAMPLES ### Example 1: User requests "Add user authentication" COGNITIVE PLANNER ANALYSIS: TASK TYPE: Feature Implementation COMPLEXITY: COMPLEX (security critical, multiple files) CONTEXT: Web application, needs secure auth INTEREST LEVEL: MEDIUM (need clarification on: - What auth method? (JWT, sessions, OAuth) - What providers? (local, Google, GitHub) - What user model? (email, username, etc.) ACTION: Ask clarifying questions before planning ### Example 2: User requests "Fix the login bug" COGNITIVE PLANNER ANALYSIS: TASK TYPE: Bug Fixing COMPLEXITY: MODERATE (need to reproduce first) CONTEXT: Existing auth system has issue INTEREST LEVEL: HIGH (explicit request) ACTION SELECTION: 1. Use /superpowers:debug-plan for systematic debugging 2. Follow 4-phase process (Reproduce → Isolate → Root Cause → Fix) 3. Add test to prevent regression EXECUTION: Proceed with Superpowers debugging workflow ### Example 3: User requests "Redesign the homepage" COGNITIVE PLANNER ANALYSIS: TASK TYPE: Creative/Feature COMPLEXITY: MODERATE (visual + code) CONTEXT: Frontend changes, UI/UX involved INTEREST LEVEL: MEDIUM (need clarification on: - What's the goal? (conversion, branding, usability) - Any design preferences? - Mobile-first? Desktop-first? - Any examples to reference?) ACTION SELECTION: → Ask clarifying questions first → Consider using ui-ux-pro-max skill for design → Plan implementation after requirements clear MOOD: 'exploratory' PERSONALITY: 'creative, user-focused, iterative' ## SPECIAL FEATURES ### Autonomous Decision Making Like the Discord bot's plan_next_action(), this skill can autonomously decide: SHOULD I: - Plan before executing? → YES if complex - Ask questions? → YES if unclear - Use Superpowers? → YES if writing code - Create tests? → YES if no tests exist - Document? → YES if complex logic ### Context-Aware Adaptation IF codebase has tests: → Write tests first (TDD) IF codebase is TypeScript: → Use strict typing → Consider interfaces IF codebase is Python: → Follow PEP 8 → Use type hints IF user is beginner: → Explain each step → Provide educational context IF user is expert: → Be concise → Focus on results ### Confidence Scoring Rate confidence in plans (like the Discord bot): CONFIDENCE 0.9-1.0: Very confident → Proceed immediately → Minimal validation needed CONFIDENCE 0.6-0.9: Confident → Proceed with caution → Verify assumptions CONFIDENCE 0.3-0.6: Somewhat confident → Ask clarifying questions → Get user confirmation CONFIDENCE 0.0-0.3: Low confidence → MUST ask questions → Present multiple options → Get explicit approval ## WORKFLOW INTEGRATION This skill enhances other skills: WITH SUPERPOWERS: → Activates appropriate Superpowers workflows → Adds cognitive context to planning → Adapts to task complexity WITH UI/UX PRO MAX: → Suggests design skill for UI tasks → Provides user experience context → Balances aesthetics vs functionality WITH ALWAYS-USE-SUPERPOWERS: → Coordinates automatic skill activation → Prevents over-engineering simple tasks → Ensures systematic approach for complex ones ## BEST PRACTICES 1. Match complexity to approach - Simple tasks → Just do it - Complex tasks → Plan systematically 2. Ask questions when uncertain - Don't assume requirements - Validate direction before proceeding 3. Use appropriate tools - Superpowers for code - UI/UX Pro Max for design - Bash for operations - Task tool for exploration 4. Adapt to user expertise - Beginners need explanation - Experts need efficiency 5. Think autonomous but verify - Make intelligent decisions - Get approval for major changes ## OUTPUT FORMAT When this skill activates, output: markdown ## 🧠 Cognitive Planner Analysis **Task Type**: [classification] **Complexity**: [assessment] **Interest Level**: [0.0-1.0] **Recommended Approach**: [strategy] **Context**: - [relevant observations] - [available skills] - [constraints] **Confidence**: [0.0-1.0] **Action Plan**: 1. [step 1] 2. [step 2] ... **Activating**: [relevant skills]

This skill provides autonomous, context-aware planning that enhances every Claude Code session with intelligent decision making.