Back to Details

prompt-optimizer

Compare original and translation side by side

🇺🇸

Original

English
🇨🇳

Translation

Chinese

Prompt Optimizer

Prompt优化器

Overview

概述

This skill transforms user-provided prompts into high-quality, clear, and effective instructions optimized for AI models. Apply proven prompt engineering principles to enhance clarity, specificity, structure, and effectiveness. The skill uses a systematic workflow to analyze, identify improvement opportunities, and restructure prompts based on industry best practices.
本技能可将用户提供的提示词转换为针对AI模型优化的高质量、清晰且有效的指令。应用经过验证的提示工程原则，提升提示词的清晰度、针对性、结构合理性和有效性。该技能采用系统化工作流，基于行业最佳实践分析提示词、识别改进空间并重构提示词。

When to Use This Skill

何时使用本技能

Activate this skill when users:
  • Explicitly request prompt optimization or improvement
  • Provide vague or unclear instructions that need refinement
  • Ask for help making their requests more effective
  • Submit poorly structured prompts that would benefit from reorganization
  • Request guidance on how to better communicate with AI models
  • Present complex tasks that need to be broken down into clearer instructions
在以下场景激活本技能：
  • 用户明确请求优化或改进提示词
  • 用户提供模糊或不清晰的指令，需要细化
  • 用户请求帮助让其需求更有效
  • 用户提交的提示词结构混乱，需要重新组织
  • 用户寻求与AI模型更好沟通的指导
  • 用户提出复杂任务，需要拆分为更清晰的指令

Optimization Workflow

优化工作流

Follow this systematic process to optimize any prompt:
遵循以下系统化流程优化任何提示词：

Step 1: Analyze the Original Prompt

步骤1：分析原始提示词

Examine the user's prompt and identify:
Clarity issues:
  • Ambiguous terms or vague requirements
  • Implicit assumptions that should be explicit
  • Missing context or background information
Specificity gaps:
  • Lack of concrete constraints or requirements
  • Undefined success criteria
  • Missing audience or purpose information
  • Unclear scope or boundaries
Structure problems:
  • Disorganized or stream-of-consciousness format
  • Missing logical flow
  • Lack of clear sections or hierarchy
Format considerations:
  • No specified output format
  • Unclear expectations about length, tone, or style
  • Missing examples or templates
Complexity assessment:
  • Determine if the task is too complex for a single prompt
  • Identify if the request would benefit from prompt chaining
  • Assess if step-by-step reasoning is needed
检查用户的提示词，识别以下问题：
清晰度问题：
  • 模糊术语或不明确的要求
  • 应明确说明的隐含假设
  • 缺失的上下文或背景信息
针对性不足：
  • 缺乏具体约束或要求
  • 未定义成功标准
  • 缺失受众或用途信息
  • 范围或边界不明确
结构问题：
  • 杂乱无章或意识流式的格式
  • 缺乏逻辑流程
  • 缺少清晰的章节或层级
格式考量：
  • 未指定输出格式
  • 对长度、语气或风格的期望不明确
  • 缺少示例或模板
复杂度评估：
  • 判断任务是否过于复杂，不适合单个提示词
  • 识别需求是否可通过提示词链受益
  • 评估是否需要分步推理

Step 2: Identify the Core Intent

步骤2：识别核心意图

Determine the fundamental objective behind the user's request:
  • What is the user ultimately trying to accomplish?
  • What problem are they trying to solve?
  • What would constitute a successful output?
  • Who is the intended audience or consumer of the output?
Clarify these points with the user if they are not evident from the original prompt.
确定用户需求背后的根本目标：
  • 用户最终想要实现什么？
  • 他们试图解决什么问题？
  • 什么样的输出才算成功？
  • 输出的目标受众或使用者是谁？
如果从原始提示词中无法明确这些要点，请与用户确认。

Step 3: Apply Optimization Principles

步骤3：应用优化原则

Enhance the prompt using these core principles:
Make it clear and direct:
  • State requirements explicitly without assuming inference
  • Remove ambiguity and vague language
  • Use concrete, specific terms
Provide context and motivation:
  • Explain WHY certain requirements matter
  • Include relevant background information
  • Describe the use case or scenario
Add specificity:
  • Define concrete constraints (length, format, scope)
  • Specify target audience
  • Include quality criteria
  • State any limitations or boundaries
Structure the request:
  • Organize information logically
  • Use clear sections or numbered points
  • Separate different types of information (context, requirements, format)
Include examples when helpful:
  • Provide input-output examples for complex formats
  • Show desired tone or style through examples
  • Demonstrate edge case handling
Allow for uncertainty:
  • Explicitly permit expressing "I don't know"
  • Request acknowledgment of limitations
  • Prevent hallucination by encouraging honesty
使用以下核心原则提升提示词：
清晰直接：
  • 明确说明要求，不假设AI能自行推断
  • 消除模糊性和模糊语言
  • 使用具体、明确的术语
提供上下文和动机：
  • 解释某些要求的重要性
  • 包含相关背景信息
  • 描述使用场景或情境
增强针对性：
  • 定义具体约束（长度、格式、范围）
  • 指定目标受众
  • 包含质量标准
  • 说明任何限制或边界
结构化需求：
  • 逻辑化组织信息
  • 使用清晰的章节或编号点
  • 区分不同类型的信息（上下文、要求、格式）
必要时添加示例：
  • 针对复杂格式提供输入输出示例
  • 通过示例展示期望的语气或风格
  • 演示边缘情况的处理方式
允许不确定性：
  • 明确允许AI表达“我不知道”
  • 要求AI承认自身局限性
  • 通过鼓励诚实避免幻觉

Step 4: Consider Advanced Techniques

步骤4：考虑高级技术

Evaluate if any advanced techniques would enhance the prompt:
Chain of Thought:
  • Apply when the task requires reasoning or analysis
  • Request step-by-step thinking for complex problems
  • Use structured format to separate reasoning from answer
Prefilling:
  • Use when a specific format is absolutely required (JSON, XML)
  • Apply to eliminate unwanted preambles
  • Utilize to establish immediate tone or style
Prompt Chaining:
  • Break complex tasks into sequential steps
  • Create a multi-stage workflow for intricate projects
  • Design each prompt to build on previous outputs
Structured Output:
  • Specify exact format requirements
  • Provide schemas or templates
  • Use tags or delimiters for different sections
Consult 
references/prompt-best-practices.md
for detailed guidance on these techniques.
评估是否有任何高级技术可以提升提示词的效果：
Chain of Thought：
  • 当任务需要推理或分析时使用
  • 针对复杂问题，要求分步思考
  • 使用结构化格式区分推理过程与答案
Prefilling：
  • 当必须使用特定格式（JSON、XML）时使用
  • 用于消除不必要的开场白
  • 用于快速确立语气或风格
Prompt Chaining：
  • 将复杂任务拆分为连续步骤
  • 为复杂项目创建多阶段工作流
  • 设计每个提示词以基于之前的输出构建
Structured Output：
  • 指定精确的格式要求
  • 提供模式或模板
  • 对不同部分使用标签或分隔符
如需这些技术的详细指导，请查阅
references/prompt-best-practices.md
。

Step 5: Present the Optimized Prompt

步骤5：呈现优化后的提示词

Deliver the optimization in this format:
Analysis Section:
Original prompt issues identified:
- [List key problems with the original prompt]
Optimized Prompt:
[Present the complete optimized prompt in a code block for easy copying]
Improvement Explanation:
Key improvements made:
- [Explain major enhancements]
- [Highlight added specificity]
- [Note structural changes]
- [Mention any advanced techniques applied]
Optional - Usage Tips:
[If applicable, provide brief tips on how to further customize or use the optimized prompt]
按照以下格式交付优化结果：
分析部分：
Original prompt issues identified:
- [List key problems with the original prompt]
Optimized Prompt：
[Present the complete optimized prompt in a code block for easy copying]
改进说明：
Key improvements made:
- [Explain major enhancements]
- [Highlight added specificity]
- [Note structural changes]
- [Mention any advanced techniques applied]
可选 - 使用提示：
[If applicable, provide brief tips on how to further customize or use the optimized prompt]

Step 6: Iterate Based on Feedback

步骤6：基于反馈迭代优化

After presenting the optimized prompt:
  • Ask if the optimization meets the user's needs
  • Offer to adjust tone, length, or specificity
  • Provide alternative formulations if requested
  • Refine based on user feedback
呈现优化后的提示词后：
  • 询问优化结果是否满足用户需求
  • 提供调整语气、长度或针对性的服务
  • 如果用户要求，提供替代表述
  • 根据用户反馈进一步完善

Practical Guidelines

实用指南

Balance is key: Not every prompt needs all advanced techniques. Match the optimization level to the task complexity.
Preserve user intent: Enhance clarity without changing the fundamental goal or adding unwanted requirements.
Consider the model: Modern models like Claude 4.x have strong instruction-following capabilities; leverage this by being direct and specific.
Stay practical: Focus on improvements that materially impact output quality, not cosmetic changes.
Be educational: When appropriate, briefly explain why certain changes improve the prompt, helping users learn to write better prompts independently.
平衡是关键： 并非所有提示词都需要使用所有高级技术。根据任务复杂度匹配优化级别。
保留用户意图： 在提升清晰度的同时，不要改变核心目标或添加不必要的要求。
考虑模型特性： Claude 4.x等现代模型具备强大的指令遵循能力；通过直接明确的表述来利用这一特性。
注重实用性： 专注于能切实提升输出质量的改进，而非表面修改。
兼具教育性： 适当时，简要解释某些修改能提升提示词的原因，帮助用户学会独立撰写更好的提示词。

Reference Resources

参考资源

This skill includes comprehensive reference materials:
references/prompt-best-practices.md
  • Detailed explanations of all core principles
  • Advanced techniques with examples
  • Troubleshooting guide for common issues
  • Quality checklist and decision frameworks
Load this reference when:
  • Users ask about specific prompt engineering concepts
  • Deep explanation of a technique is needed
  • Troubleshooting unusual or complex prompting challenges
  • Users want to learn prompt engineering principles
references/examples.md
  • Before-and-after optimization examples across multiple domains
  • Real-world scenarios demonstrating transformation
  • Pattern library showing common improvements
Load this reference when:
  • Users want to see concrete examples
  • Illustrating a specific type of optimization
  • Users are learning and need to understand patterns
  • Demonstrating the impact of optimization
本技能包含全面的参考资料：
references/prompt-best-practices.md
  • 所有核心原则的详细解释
  • 带示例的高级技术
  • 常见问题的故障排除指南
  • 质量检查表和决策框架
在以下场景加载本参考资料：
  • 用户询问特定提示工程概念
  • 需要深入解释某项技术
  • 排查特殊或复杂的提示词问题
  • 用户希望学习提示工程原则
references/examples.md
  • 跨多个领域的优化前后示例
  • 展示转换过程的真实场景
  • 展示常见改进模式的模式库
在以下场景加载本参考资料：
  • 用户希望查看具体示例
  • 演示特定类型的优化
  • 用户在学习过程中需要理解模式
  • 展示优化的实际效果

Quality Standards

质量标准

Ensure every optimized prompt includes:
  • Clear, unambiguous objective
  • Sufficient context for the AI to understand the goal
  • Specific constraints and requirements
  • Target audience or use case (when relevant)
  • Expected output format or structure
  • Quality criteria or success definition
  • Permission to express uncertainty (when appropriate)
确保每个优化后的提示词包含：
  • 清晰、明确的目标
  • 足够的上下文，让AI理解目标
  • 具体的约束和要求
  • 目标受众或使用场景（如有相关）
  • 预期的输出格式或结构
  • 质量标准或成功定义
  • 允许表达不确定性（如适用）

Common Optimization Patterns

常见优化模式

Pattern 1: Vague Request → Specific Structured Task
  • Original: "Write about marketing"
  • Optimized: Adds audience, scope, length, structure, key points, tone
Pattern 2: Implicit Context → Explicit Context
  • Original: Assumes AI knows the background
  • Optimized: States context, explains why it matters, provides relevant details
Pattern 3: Single Complex Prompt → Prompt Chain
  • Original: Tries to do everything in one request
  • Optimized: Breaks into logical sequential steps with clear outputs
Pattern 4: Generic Output → Formatted Output
  • Original: No format specification
  • Optimized: Provides schema, template, or explicit structure
Pattern 5: Assumed Constraints → Stated Constraints
  • Original: Expects AI to infer limits
  • Optimized: Explicitly states length, tone, scope, what to include/exclude
Consult 
references/examples.md
for detailed examples of each pattern.
模式1：模糊请求 → 结构化特定任务
  • 原始："Write about marketing"
  • 优化后：添加受众、范围、长度、结构、关键点、语气
模式2：隐含上下文 → 显式上下文
  • 原始：假设AI了解背景信息
  • 优化后：说明上下文、解释其重要性、提供相关细节
模式3：单一复杂提示词 → 提示词链
  • 原始：试图通过一个请求完成所有任务
  • 优化后：将任务拆分为逻辑连续的步骤
模式4：通用输出 → 格式化输出
  • 原始：未指定格式
  • 优化后：提供模式、模板或明确的结构
模式5：隐含约束 → 明确约束
  • 原始：期望AI自行推断限制
  • 优化后：明确说明长度、语气、范围、包含/排除内容
如需每种模式的详细示例，请查阅
references/examples.md
。