metathink
Metathink - Optimize Claude Code usage through metacognitive strategies. Use extended thinking, context engineering, task decomposition, and chain-of-thought prompting. Use when planning complex tasks, debugging difficult issues, making architectural decisions, or improving AI collaboration effectiveness.
npx skills add majiayu000/claude-skill-registry --skill metathink --agent claude-code
Same command for any agent — swap --agent for codex, cursor, copilot.
Weekly change comes from our own snapshots, not the repository page — it measures attention, not adoption.
What it does
Metathink provides a strategic framework to optimize Claude Code effectiveness by encouraging metacognitive awareness during coding tasks. It covers when to use extended thinking, how to construct prompts that reveal reasoning, and how to manage context and task decomposition to improve collaboration and planning before implementation.
How it works
- Trigger extended thinking through explicit prompts such as "think", "think hard", and "think harder" to allocate larger reasoning budgets.
- Use a metacognitive checklist (Reflect, Plan, Prompt, Monitor, Evaluate, Iterate) before and during tasks to guide reasoning and decision-making.
- Follow a core workflow that includes assessing task complexity, selecting appropriate agents, and employing context management strategies to load knowledge just-in-time.
- Employ plan mode and debounced reasoning visibility via Ctrl+O to observe Claude's thinking, and use markers like /clear to reset context between tasks.
- Utilize multiple prompts for explicit reasoning, trade-off analyses, and exploration before coding.
- Apply a structured task-decomposition approach to break complex work into phases and delegate to specialized agents when needed (e.g., database-specialist, db-performance-agent).
When to use it
- Complex architectural decisions
- Difficult debugging
- Large refactoring tasks
- Learning new patterns
- Session optimization
- Context management
- Quality improvement
What it can touch
- Tools and commands mentioned include: /clear, Ctrl+O, /help. The skill also references delegating to agents like the database-specialist or db-performance-agent, and using prompts that request reasoning and plan mode before coding.
Caveats
- The framework advocates explicit reasoning prompts and monitoring but does not guarantee correctness of outcomes.
- It emphasizes the use of metacognitive prompts and plan-mode workflows, which may increase prompt length and reasoning time.
- It requires user discipline to engage in the Reflect/Plan/Prompt/Monitor/Evaluate/Iterate loop and to reset context between tasks with /clear.
# Metathink: Metacognitive Reasoning with Claude Code Strategic framework for optimizing Claude Code effectiveness through metacognitive awareness—thinking about how you and Claude think together. > **Metathink** = Thinking about thinking while coding ## When to Use This Skill - **Complex architectural decisions** - Need deep reasoning about trade-offs - **Difficult debugging** - Issues requiring multi-step investigation - **Large refactoring tasks** - Planning before implementation critical - **Learning new patterns** - Understanding "why" not just "what" - **Session optimization** - Improving collaboration effectiveness - **Context management** - Maintaining clean, relevant context - **Quality improvement** - Moving beyond "metacognitive laziness" ## Quick Reference ### Trigger Extended Thinking ```bash # In your prompts, use trigger words for deeper reasoning: "think" # ~4,000 token thinking budget "think hard" # ~10,000 tokens "think harder" # ~31,999 tokens (maximum) "ultrathink" # Maximum reasoning budget # View Claude's internal reasoning Ctrl+O # Toggle verbose mode (gray italic text) ``` ### Essential Commands ```bash /clear # Reset context between distinct tasks Ctrl+O #
- When to Use This Skill
- Quick Reference
- Trigger Extended Thinking
- Essential Commands
- Quick Metacognitive Checklist
- Core Workflow
- 1. Metacognitive Task Assessment
- 2. Strategic Prompt Construction
- 3. Context Management
- 4. Task Decomposition Strategy
- 5. Chain-of-Thought Prompting
- 6. Monitoring Claude's Reasoning
- 7. Avoiding Metacognitive Laziness
- Common Patterns
In your prompts, use trigger words for deeper reasoning: View Claude's internal reasoning Clear context between tasks Use skills for just-in-time knowledge Delegate to specialized agents Load specific context files when needed
What does the metathink skill do?
Metathink - Optimize Claude Code usage through metacognitive strategies. Use extended thinking, context engineering, task decomposition, and chain-of-thought prompting. Use when planning complex tasks, debugging difficult issues, making architectural decisions, or improving AI collaboration effectiveness.
How do I install it?
Run `npx skills add majiayu000/claude-skill-registry --skill metathink --agent claude-code` — it drops the skill into your project so the agent can pick it up. Swap the --agent value for codex, cursor or copilot if you use one of those.
Where does this skill come from?
From majiayu000/claude-skill-registry, a repository with 534 stars. We read it straight from the repository tree rather than a submitted listing, so what you see here is what is actually published.
Is a popular skill a good skill?
Not necessarily. Stars measure attention, not adoption — a repository can trend for a week and be abandoned. That is why we show the weekly change from our own snapshots next to the total, instead of a single flattering number.