strategist

**[TIER 1: HIGH FIDELITY] when you need comprehensive delegated analysis** with deep exploration and framework guidance. Guide framework selection and delegate to strategist subagent for thorough analysis.

[Prompts] Create single prompts, prompt chains (sequential workflows), or meta-prompts (prompts that generate other prompts). Examples: Create a prompt for security code review | Create a prompt chain for competitive analysis | Create a meta-prompt that generates task-specific prompts.

Delegate to prompt-engineer for prompt optimization and design. Examples: Optimize this prompt for better results | Design a prompt for code review | Fix this prompt that's not working.

[Prompts] Refine existing prompts through fast edit, improvement, or full refactoring using meta-prompt optimization. Examples: Fix the output format in my code-review prompt | Improve my analysis prompt with better examples | My prompt is producing inconsistent results, refactor it completely.

[Execution] Execute a saved prompt with proper context loading and structured output handling. <example> Context: User wants to run a saved prompt user: "Run my code-review prompt on this file" assistant: "I will use /run-prompt to execute the saved prompt." </example>

Guide framework selection and apply thinking frameworks to analyze problems in the current conversation context using the thinking-frameworks skill resources. <example> Context: User wants structured thinking user: "Think through this problem systematically" assistant: "I'll guide you through framework selection and analysis." </example> <example> Context: Interactive problem solving user: "Help me think about this decision" assistant: "I'll use the think command to apply structured frameworks." </example> <example> Context: Foreground framework application user: "Apply First Principles thinking to this" assistant: "I'll guide you through applying the framework interactively." </example>