AI Prompt Engineering Cheatsheet: Essential Action Verbs for Better Results

Abstract

Effective AI prompt engineering requires precise action verbs to eliminate ambiguity and improve output quality. This cheatsheet categorizes 60+ essential verbs across eight functional domains with comparative examples demonstrating improved prompt construction.

Methodology

Action verbs are categorized by cognitive and operational functions. Each category includes weak versus strong prompt examples demonstrating measurable improvement in specificity and output quality.

Verb Categories and Applications

Analysis & Understanding

Function: Information examination and interpretation Implementation: Replace vague requests with specific analytical tasks

Weak: Tell me about this code
Strong: Analyze this Python function for security vulnerabilities and performance bottlenecks

Creation & Generation

Function: Novel content and solution development Implementation: Specify deliverable type, constraints, and success criteria

Weak: Make a website layout
Strong: Design a responsive landing page layout for SaaS product with hero section, feature highlights, and CTA buttons

Explanation & Teaching

Function: Knowledge transfer and conceptual clarification Implementation: Define audience expertise level and learning objectives

Weak: What is Docker?
Strong: Explain Docker containerization to a developer familiar with VMs, including key benefits and basic commands

Problem-Solving

Function: Issue resolution and optimization Implementation: Identify specific problems and define optimization targets

Weak: This code isn't working
Strong: Debug this React component causing infinite re-renders and optimize for performance

Information Processing

Function: Data transformation and content manipulation Implementation: Specify input format, desired output, and transformation parameters

Weak: Change this text
Strong: Rewrite technical blog post for non-technical audience while preserving key concepts

Reasoning & Logic

Function: Evidence-based analysis and conclusion generation Implementation: Request logical frameworks and justification criteria

Weak: What do you think will happen?
Strong: Predict microservices implementation outcomes based on current team size and technical constraints

Creative Tasks

Function: Original ideation and innovative solution development Implementation: Define creative constraints and evaluation criteria

Weak: Give me some ideas
Strong: Brainstorm 10 innovative features for project management tool targeting remote software teams

Research & Discovery

Function: Information gathering and synthesis Implementation: Specify scope, sources, and synthesis requirements

Weak: Find information about React
Strong: Research React 18 features and compile comparison with React 17 focusing on performance improvements

Optimization Strategies

Verb Combination: Layer multiple verbs for complex tasks

Analyze user feedback data, categorize issues by severity, generate actionable recommendations

Context Specification: Include technical constraints and success metrics

Optimize database query for high-traffic e-commerce site (10M+ daily users) focusing on read performance

Quick Reference: Complete Verb Taxonomy

Category	Primary Function	Complete Verb Set
Analysis & Understanding	Information examination	analyze, evaluate, examine, interpret, compare, contrast, critique, investigate, diagnose, identify, distinguish, categorize, classify, assess
Creation & Generation	Content development	create, generate, write, compose, draft, develop, design, formulate, devise, build, synthesize, compile, construct
Explanation & Teaching	Knowledge transfer	explain, describe, clarify, illustrate, demonstrate, teach, guide, walk through, break down, simplify, elaborate, define, outline, summarize
Problem-Solving	Issue resolution	solve, troubleshoot, debug, fix, resolve, optimize, improve, refine, streamline, transform, convert, adapt, modify, refactor
Information Processing	Data manipulation	extract, organize, structure, format, condense, expand, translate, paraphrase, rewrite, edit, revise, fact-check, compile
Reasoning & Logic	Evidence-based analysis	reason, deduce, infer, conclude, predict, calculate, determine, validate, verify, justify, argue, reflect, chain
Creative Tasks	Innovation and ideation	brainstorm, imagine, innovate, conceive, invent, reimagine, roleplay, simulate, model, visualize, prototype, storyboard
Research & Discovery	Information gathering	research, investigate, explore, search, find, discover, trace, track, locate, gather, curate, contextualize
Process & Coordination	Task management	iterate, prioritize, orchestrate, systematize, decompose, sequence, coordinate, schedule, delegate

Claude-Specific Optimizations

Leveraging Claude's Strengths:

• Parallel processing: Use "orchestrate" for multi-tool coordination
• Extended context: Apply "contextualize" for large document analysis
• Structured thinking: Employ "chain" for step-by-step reasoning
• Reflection capabilities: Utilize "reflect" for self-assessment and iteration
• Format control: Implement "structure" and "format" with XML tags

Claude-Optimized Verb Combinations:

Analyze this codebase, then decompose the issues by priority, and systematize solutions for each component

Error Prevention

Common Failures:

• Vague verb usage ("help me with")
• Multiple conflicting objectives
• Insufficient context specification
• Absence of success criteria

Mitigation: Apply specific verbs with defined parameters, constraints, and measurable outcomes.

Conclusion

Precise verb selection significantly improves AI prompt effectiveness. Implementation of this taxonomy enhances output quality through reduced ambiguity and improved task specification. Claude 4 models respond well to clear, explicit instructions and benefit from structured, specific prompting approaches.