sequential-thinking

Structure complex problem-solving into sequential thoughts with revision tracking and branching.

1|1|Updated Nov 9, 2025
One-click install
npx skills add https://github.com/tianguzhe/zsh-config --skill sequential-thinking-tianguzhe
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: sequential-thinking
Source: https://github.com/tianguzhe/zsh-config/tree/main/claude/skills/sequential-thinking
Command: npx skills add https://github.com/tianguzhe/zsh-config --skill sequential-thinking-tianguzhe

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill enables AI agents to solve complex problems by enforcing structured, sequential thinking with built-in revision and branching.

Core Features & Use Cases

  • Iterative reasoning: Break complex problems into sequential thoughts.
  • Revision tracking: Update conclusions when new evidence emerges.
  • Branch exploration: Evaluate multiple alternatives in parallel.
  • Maintained context: Preserve the reasoning chain across steps.
  • Use Case: Planning architecture, debugging, and design decisions requiring phased analysis.

Quick Start

Install the Sequential Thinking MCP server, copy this skill into your project's .claude/skills directory, and restart Claude to activate. Then invoke the skill with a prompt such as: "Let's think step-by-step using sequential thinking."

Frequently Asked Questions about sequential-thinking

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
What is structured sequential thinking for complex problem-solving?

Structured sequential thinking breaks complex problems into iterative reasoning steps, preserving context across stages. It supports revision tracking to update conclusions when new evidence emerges and branching to explore multiple design alternatives in parallel.

How do I break down architecture planning into iterative reasoning steps?

Architecture planning can be broken down by enforcing sequential thinking that maintains the reasoning chain across steps. This allows dynamic scope adjustments, phased analysis, and branch exploration to evaluate multiple architectural alternatives in parallel.

Can I revise previous conclusions when new evidence emerges during multi-step analysis?

Yes, revision tracking allows you to update conclusions when new evidence emerges during multi-step analysis. The mechanism preserves the reasoning chain while enabling dynamic scope adjustments to incorporate new information without losing context.

Does sequential thinking support branching to evaluate multiple design alternatives?

Yes, sequential thinking supports branching to evaluate multiple design alternatives in parallel. This allows you to explore different architectural decisions or debugging approaches simultaneously while maintaining the full reasoning context across each branch.

Do I need the Sequential Thinking MCP server to activate structured reasoning?

Yes, the Sequential Thinking MCP server is required to activate structured reasoning. You need to install the server, copy the skill into your project's .claude/skills directory, and restart Claude to enable iterative problem-solving with revision tracking.

When should I use branch exploration instead of linear step-by-step reasoning?

Branch exploration should be used when multiple viable alternatives exist for design decisions or architectural choices. Linear sequential reasoning is sufficient for straightforward debugging, while branching enables parallel evaluation of competing approaches with maintained context.