Thought Based Reasoning

This is a comprehensive reference guide for getting Claude to show its work on complex problems. It covers eight reasoning techniques with actual prompt templates, decision matrices showing accuracy gains, and implementation examples. The core idea: adding "Let's think step by step" can boost accuracy 20-60% on reasoning tasks, and the guide walks through when to use zero-shot CoT versus few-shot versus heavier approaches like Tree of Thoughts for problems requiring backtracking. Each technique includes research citations, strengths, limitations, and concrete code snippets. Most useful when you're hitting a wall with simple prompts on math, logic, or multi-step problems and need a systematic approach backed by actual research numbers.

skills.sh installs: 1.4K
GitHub stars: 1.8K
last updated: Aug 21, 2026

View on GitHub

Install

$ npx -y skills add neolabhq/context-engineering-kit --skill thought-based-reasoning --agent claude-code

Installs into .claude/skills of the current project.

Files

Thought-Based Reasoning Techniques for LLMs

Overview

Chain-of-Thought (CoT) prompting and its variants encourage LLMs to generate intermediate reasoning steps before arriving at a final answer, significantly improving performance on complex reasoning tasks. These techniques transform how models approach problems by making implicit reasoning explicit.

Quick Reference

Technique	When to Use	Complexity	Accuracy Gain
Zero-shot CoT	Quick reasoning, no examples available	Low	+20-60%
Few-shot CoT	Have good examples, consistent format needed	Medium	+30-70%
Self-Consistency	High-stakes decisions, need confidence	Medium	+10-20% over CoT
Tree of Thoughts	Complex problems requiring exploration	High	+50-70% on hard tasks
Least-to-Most	Multi-step problems with subproblems	Medium	+30-80%
ReAct	Tasks requiring external information	Medium	+15-35%
PAL	Mathematical/computational problems	Medium	+10-15%
Reflexion	Iterative improvement, learning from errors	High	+10-20%