Token导航 LogoToken导航

Agent Skill · 开发

smart-prompt-generator(智能提示生成器)

用于辅助提示词、系统指令、Agent 行为约束和工作流模板的整理。它适合让 Agent 规范任务边界、统一输出格式、拆分操作步骤或优化提示词可复用性。使用时需要保留真实业务约束,不要把示例当硬规则;涉及自动执行、外部工具或高风险操作时,应在提示词中明确确认步骤、权限边界和失败处理方式。

smart-prompt-generator 是开发类 Agent Skill,可用宿主为 OpenClaw,通过ClawHub获取,权限为需要联网;许可证为 MIT-0,第三方扫描已通过,按安装量在开发类的 21,746 个 Skill 里排第 4,128。

总安装 8,985周安装 382GitHub Stars 公开资料未说明下载 3,148核验 2026-05-01

安装说明

本站只整理中文说明,不托管安装包,也不代用户安装。
GitHub
许可证
MIT-0
版本
1.0.0

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:smart-prompt-generator(智能提示生成器)
来源仓库:https://github.com/helal-muneer/smart-prompt-generator
安装命令:
openclaw skills install smart-prompt-generator
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 OpenClaw 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

ClawHubOpenClaw
openclaw skills install smart-prompt-generator

这个 Skill 是做什么的

开发类 Skill 面向写代码、改代码和排查问题,一般会读写仓库文件,部分还会执行构建或测试命令。

智能提示词生成器帮助构建高质量、可复用的 AI 任务指令模板。适用于创意写作、编程、商业分析等多种场景的提示工程优化。支持 Flutter、React 等技术栈专项提示生成,提升开发效率。在 OpenClaw 中规划复杂任务前调用以标准化输入格式。使用时需保留真实业务约束,避免将示例当作硬性规则执行。

上手步骤

4 步
  1. 1

    确认自己的宿主工具

    公开资料标注支持 OpenClaw;先确认自己用的是其中之一,再按对应工具的 Skills 目录放置。

  2. 2

    拿到安装命令

    可以直接复制的命令是 openclaw skills install smart-prompt-generator;命令里的占位符换成自己的路径或密钥再执行。

  3. 3

    安装后重启宿主

    命令执行完要完全退出宿主工具再打开,让新装的 Skill 被重新扫描到,否则列表里可能看不到。

  4. 4

    按最小权限试用

    它需要需要联网,先在测试目录或测试账号里跑,确认没问题再放到正式项目。

常见问题

6 条

smart-prompt-generator 怎么安装?

在终端执行 openclaw skills install smart-prompt-generator。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

smart-prompt-generator 支持哪些 AI 工具?

公开资料标注支持 OpenClaw。其他工具只要能读取本地 Skill 目录,一般也能手动放进去,但要自己确认触发方式是否一致。

smart-prompt-generator 需要什么权限?

权限级别是需要联网。该 Skill 可能需要联网访问来源站点、仓库或外部 API;具体网络访问范围需要结合源码和 README 复核。

smart-prompt-generator 是免费的吗?

来源标注的许可证是 MIT-0。本站不代收任何费用,也不托管安装包;能否商用、要不要署名,以许可证原文为准。

smart-prompt-generator 用的人多吗?

公开数据是累计安装 8,985 次,近一周安装 382 次,下载 3,148 次。这些数字来自第三方目录统计,只能当作热度参考,不代表质量或安全性。

smart-prompt-generator 安全吗?

第三方扫描结果为通过,但这只覆盖已知的静态检查项,不能替代自己读一遍源码和安装命令。

值得选的地方和要先想清楚的地方

值得选的地方

  • 专注 OpenClaw 一种宿主,触发方式和使用习惯不用来回切换。
  • 许可证明确为 MIT-0,商用和分发边界好判断。
  • 第三方扫描已通过,常见的高风险行为没有命中。

先想清楚的地方

  • 第三方目录整理,版本更新和行为细节以来源仓库为准。

适不适合你

更适合

  • 已经在用 OpenClaw,想直接扩展它能力的人。
  • 需要处理开发类任务、又不想每次都手写提示词的人。
  • 需要许可证清楚、并且有第三方扫描结果的团队项目。

不太适合

  • 需要厂商级技术支持和 SLA 的生产系统。

开发类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
skill-vetter

技能审查员 · OpenClaw

8,779,018986操作浏览器MIT-0通过
ontology

本体论 · OpenClaw

6,636,960576只读MIT-0通过
entra-app-registration

内部应用程序注册 · Codex、Claude

6,631,872764接触敏感数据未明确通过
entra-app-registration

内部应用程序注册 · Codex、Claude

3,757,680201接触敏感数据未明确通过
nano-banana-pro

纳米香蕉专业版 · OpenClaw

3,526,898364接触敏感数据MIT-0有提醒
smart-prompt-generator

智能提示生成器 · 本页

8,985—需要联网MIT-0通过

按安装量取同分类前几名;数字来自第三方目录统计,只作热度参考,不代表质量结论。

SKILL.md 原文

name
prompt-generator
description
Generate high-quality, customizable AI prompts for various use cases including creative writing, problem-solving, education, business, programming, Flutter development, and game development. Use when a user asks for help creating prompts, wants to improve their AI interactions, or needs structured prompt templates for specific tasks.

AI Prompt Generator

Overview

Generate high-quality, customizable AI prompts that help users get better results from AI tools. This skill provides structured templates, best practices, and prompt generation for various use cases including Flutter app development and game development.

Core Features

  • Multiple Prompt Categories: Creative, Problem-Solving, Educational, Business, Programming, Flutter Development, Game Development
  • Tone & Style Customization: Formal, Casual, Technical, Creative, Professional
  • Length & Complexity Control: Short, Medium, Detailed | Beginner, Intermediate, Advanced
  • Context Enhancement: Add background, constraints, and specific requirements
  • Template Library: Pre-built templates with customization options
  • Real-time Guidance: Best practices and tips for effective prompts

Usage

Basic Usage

  1. Select Prompt Type: Choose from available categories
  2. Define Your Goal: What do you want the AI to help with?
  3. Customize: Adjust tone, length, complexity
  4. Add Context: Include relevant background information
  5. Generate: Get your structured prompt
  6. Use: Copy and paste into your AI tool

Advanced Usage

  1. Template System: Save and load custom templates
  2. Iterative Refinement: Improve prompts based on results
  3. Multi-step Prompts: Create complex, multi-stage prompts
  4. Constraint Addition: Add specific requirements and limitations
  5. Output Formatting: Define how you want the response structured

Prompt Categories

1. Creative Writing Prompts

Structure:

Role: [creative writer/editor]
Task: [specific writing task]
Context: [background, setting, characters]
Constraints: [word count, style, tone]
Output: [expected format]

Example:

You are a creative fiction writer. Write a detailed short story about an astronaut's 
experience on a solo mission. Include emotional challenges, a turning point in the 
plot, and a meaningful ending. Target length: 1500 words. Tone: Reflective and 
introspective.

2. Problem-Solving Prompts

Structure:

Context: [situation/background]
Problem: [specific issue]
Constraints: [limitations, requirements]
Desired Output: [solution format]
Success Criteria: [how to measure success]

Example:

I'm experiencing [specific technical issue] in [context/environment]. The problem 
occurs when [specific conditions]. I've tried [previous attempts]. Constraints: 
[cannot change X, must work with Y]. Provide step-by-step troubleshooting with 
explanations for each step.

3. Educational Prompts

Structure:

Topic: [subject matter]
Audience: [who is learning]
Learning Objective: [what they should understand]
Teaching Method: [explanation style]
Assessment: [how to verify understanding]

Example:

Explain [complex concept] to [beginners/intermediate/advanced learners] using 
[analogies/examples/step-by-step approach]. Include practical examples and 
address common misconceptions. Provide a brief quiz at the end to check 
understanding.

4. Flutter App Development Prompts

Structure:

App Type: [what kind of app]
Features: [required functionality]
State Management: [Provider/Riverpod/Bloc/etc.]
Architecture: [MVVM/Clean Architecture/etc.]
Platforms: [iOS/Android/Web/Desktop]

Templates:

App Architecture Setup
Create a Flutter app architecture for [app type] with the following requirements:
- State Management: [Provider/Riverpod/Bloc/GetX]
- Architecture Pattern: [MVVM/Clean Architecture/MVC]
- Key Features: [list main features]
- Target Platforms: [iOS/Android/Web/Desktop]
- Data Persistence: [Hive/SharedPreferences/SQLite/Firebase]
- API Integration: [REST/GraphQL/Firebase]

Provide:
1. Folder structure
2. Key files and their responsibilities
3. Dependency injection setup
4. Navigation approach
5. Error handling strategy
Feature Implementation
Implement [specific feature] in Flutter with these requirements:
- Feature Description: [detailed description]
- UI/UX Requirements: [design specifications]
- State Management: [how state will be managed]
- Data Flow: [how data moves through the app]
- Error Handling: [how errors are handled]
- Testing: [unit/widget/integration tests needed]

Include:
1. Complete code implementation
2. State management setup
3. UI components
4. Business logic
5. Error handling
6. Example usage
Widget Creation
Create a custom Flutter widget for [purpose] with these specifications:
- Widget Type: [stateless/stateful]
- Props/Parameters: [required and optional parameters]
- Styling: [theming and customization options]
- Accessibility: [semantic labels, screen reader support]
- Performance: [optimization considerations]

Provide:
1. Widget code with documentation
2. Example usage
3. Customization options
4. Performance optimizations
5. Accessibility features
State Management Setup
Set up state management for [app feature] using [Provider/Riverpod/Bloc]:

Requirements:
- State Type: [simple/complex/nested]
- Updates: [how state changes]
- Persistence: [local storage needs]
- Performance: [optimization requirements]

Include:
1. State model/class definition
2. State management setup
3. Provider/Bloc/Cubit implementation
4. Consumer widget examples
5. Testing approach
API Integration
Integrate [API name] into Flutter app with these requirements:
- API Type: [REST/GraphQL/Firebase]
- Endpoints: [list required endpoints]
- Authentication: [method]
- Error Handling: [strategy]
- Caching: [requirements]
- Offline Support: [needs]

Provide:
1. API client setup
2. Model classes
3. Repository pattern implementation
4. Error handling
5. Caching strategy
6. Example usage

5. Flutter Game Development Prompts

Structure:

Game Type: [2D/2.5D/puzzle/platformer/etc.]
Engine: [Flame/Forge2D/custom]
Core Mechanics: [main gameplay elements]
Art Style: [visual approach]
Platforms: [mobile/web/desktop]

Templates:

Game Architecture Setup
Design a Flutter game architecture for [game type] using [Flame/Forge2D]:

Requirements:
- Game Type: [puzzle/platformer/RPG/etc.]
- Core Mechanics: [list main gameplay elements]
- Save System: [progress saving needs]
- Monetization: [ads/IAP/premium]
- Platforms: [mobile/web/desktop]

Include:
1. Game loop structure
2. Scene/screen management
3. Component system
4. State management
5. Asset management
6. Save/load system
7. Performance optimizations
Game Mechanic Implementation
Implement [specific game mechanic] in Flutter using [Flame/Forge2D]:

Mechanic Details:
- Type: [movement/collision/scoring/etc.]
- Player Interaction: [input methods]
- Visual Feedback: [animations/effects]
- Audio: [sound effects/music]

Provide:
1. Core mechanic code
2. Component setup
3. Collision detection (if needed)
4. Animation system
5. Input handling
6. Example usage
7. Testing approach
Character/Entity System
Create a character/entity system for [game type] with:

Requirements:
- Entity Types: [player/enemies/NPCs/objects]
- Behaviors: [AI/movement/interaction]
- Attributes: [health/score/abilities]
- Rendering: [sprites/animations]

Include:
1. Base entity class
2. Component system
3. Behavior trees/state machines
4. Animation system
5. Collision system
6. Example entities
7. Performance optimizations
Level/World Design
Design a level/world system for [game type]:

Requirements:
- Level Structure: [linear/open-world/procedural]
- Progression: [difficulty scaling/unlocks]
- Save Points: [checkpoints/persistence]
- Assets: [tilemaps/3D models/etc.]

Provide:
1. Level data structure
2. Level loading system
3. Procedural generation (if applicable)
4. Progression system
5. Asset management
6. Example levels
7. Testing approach
Game UI/UX
Design game UI/UX for [game type] with these screens:

Screens Needed:
- Main Menu: [options/features]
- HUD: [score/health/controls]
- Pause Menu: [options/controls]
- Game Over: [stats/restart/quit]

Include:
1. UI component structure
2. State management for UI
3. Animation/transitions
4. Responsive design
5. Accessibility features
6. Example implementations
Performance Optimization
Optimize Flutter game performance for [platform]:

Current Issues:
- [specific performance problems]
- [bottlenecks identified]

Target Performance:
- Frame Rate: [target FPS]
- Memory: [memory budget]
- Load Times: [acceptable load times]

Provide:
1. Profiling approach
2. Optimization strategies
3. Code improvements
4. Asset optimization
5. Memory management
6. Testing methodology

6. Business & Marketing Prompts

Templates:

Marketing Campaign
Create a marketing campaign for [product/service] targeting [audience]:

Campaign Goals: [awareness/conversion/engagement]
Target Audience: [demographics, interests]
Key Messages: [unique value propositions]
Channels: [social media/email/ads]
Budget: [constraints]
Timeline: [duration]

Include:
1. Campaign strategy
2. Content calendar
3. Messaging framework
4. Channel strategy
5. KPIs and measurement
6. Example content pieces
Business Strategy
Develop a business strategy for [company/product]:

Current Situation: [market position, challenges]
Goals: [short-term and long-term objectives]
Resources: [available assets and constraints]
Market: [industry trends, competition]

Provide:
1. Strategic analysis
2. Competitive positioning
3. Growth strategy
4. Implementation roadmap
5. Risk mitigation
6. Success metrics

Best Practices

1. Start Clear

  • Define your exact goal
  • Identify your audience
  • Specify desired output format
  • Include relevant context

2. Be Specific

  • Use concrete examples
  • Define technical terms
  • Specify scope and depth
  • Include constraints

3. Add Context

  • Provide background information
  • Include relevant constraints
  • Mention preferred style/approach
  • Define success criteria

4. Iterate

  • Test generated prompts
  • Refine based on results
  • Save successful versions
  • Build on what works

5. Quality Check

  • Review for clarity
  • Ensure completeness
  • Verify alignment with goals
  • Check for ambiguity

Prompt Enhancement Framework

Context Enhancement

[Base Prompt]
Context: [Background information]
Constraints: [Limitations and requirements]
Examples: [Sample inputs/outputs]
Format: [Desired output structure]

Constraint Addition

[Base Prompt]
Requirements:
- Must include [specific elements]
- Must avoid [specific elements]
- Must be [length/format]
- Must consider [specific aspects]

Output Formatting

[Base Prompt]
Output Format:
- Structure: [paragraph/list/table/etc.]
- Length: [word count/sections]
- Style: [tone/voice]
- Include: [specific sections/elements]

Common Prompt Patterns

Role-Based Prompts

You are a [specific role] with expertise in [domain]. 
[Task description] considering [specific context].
Provide [output format] that includes [specific elements].

Step-by-Step Prompts

[Task description]. Break this down into:
1. [First step]
2. [Second step]
3. [Third step]
For each step, provide [specific details].

Example-Driven Prompts

[Task description]. Follow this pattern:
Example 1: [input] → [output]
Example 2: [input] → [output]
Now apply this to: [new input]

Comparison Prompts

Compare [Option A] and [Option B] for [specific use case].
Consider: [criteria 1], [criteria 2], [criteria 3].
Provide recommendation with justification.

Prompt Refinement Techniques

Iterative Refinement

  1. Start with basic prompt
  2. Test with AI tool
  3. Identify gaps/issues
  4. Add missing context
  5. Refine language
  6. Test again
  7. Repeat until satisfied

Quality Indicators

  • ✅ Clear and unambiguous
  • ✅ Specific and focused
  • ✅ Includes relevant context
  • ✅ Defines success criteria
  • ✅ Provides examples (if helpful)
  • ✅ Specifies output format

Common Pitfalls to Avoid

  • ❌ Too vague or broad
  • ❌ Missing context
  • ❌ Conflicting instructions
  • ❌ Unrealistic expectations
  • ❌ No success criteria
  • ❌ Ignoring AI limitations

Template Customization Guide

For Beginners

  1. Use pre-built templates as-is
  2. Focus on filling in placeholders
  3. Keep prompts simple and direct
  4. Test and iterate frequently

For Intermediate Users

  1. Customize templates to your needs
  2. Combine multiple prompt patterns
  3. Add domain-specific context
  4. Create your own templates

For Advanced Users

  1. Build complex, multi-step prompts
  2. Create prompt libraries for reuse
  3. Develop domain-specific frameworks
  4. Optimize for specific AI models

Usage Examples

Example 1: Flutter App Feature

Input: "I need to add user authentication to my Flutter app"

Generated Prompt:
"Implement user authentication in a Flutter app with the following requirements:
- Authentication Methods: Email/password, Google Sign-In, Apple Sign-In
- State Management: Riverpod for auth state
- Persistence: Keep user logged in across app restarts
- Security: Secure token storage using flutter_secure_storage
- Error Handling: User-friendly error messages
- UI: Login screen with form validation

Provide:
1. Complete authentication service implementation
2. Riverpod providers setup
3. UI implementation with form validation
4. Error handling strategy
5. Testing approach
6. Example usage in widgets"

Example 2: Flutter Game Mechanic

Input: "I want to add a jumping mechanic to my platformer game"

Generated Prompt:
"Implement a jumping mechanic for a 2D platformer game in Flutter using Flame:

Requirements:
- Jump Types: Single jump, variable height (based on button press duration)
- Physics: Realistic gravity and jump arc
- Input: Touch/screen tap for jump
- Visual Feedback: Jump animation, dust particles on landing
- Audio: Jump sound effect, landing sound
- Controls: Variable jump height based on press duration

Include:
1. Player component with jump logic
2. Physics implementation (gravity, velocity)
3. Input handling for variable jump
4. Animation system for jump states
5. Particle effects for landing
6. Audio integration
7. Example usage
8. Performance optimizations"

Example 3: Business Strategy

Input: "Help me create a marketing strategy for my new mobile app"

Generated Prompt:
"Develop a comprehensive marketing strategy for a new mobile app launch:

App Details:
- Type: [productivity/game/social/etc.]
- Target Audience: [demographics, behaviors]
- Unique Value: [key differentiators]
- Platforms: [iOS/Android/both]

Marketing Goals:
- User Acquisition: [target number of users]
- Engagement: [retention goals]
- Monetization: [revenue model]

Budget: [total marketing budget]
Timeline: [pre-launch, launch, post-launch phases]

Provide:
1. Market analysis and positioning
2. Target audience personas
3. Marketing channel strategy
4. Content marketing plan
5. User acquisition tactics
6. Retention and engagement strategies
7. Budget allocation
8. KPIs and measurement framework
9. Launch timeline
10. Risk mitigation strategies"

Troubleshooting

Issue: Generated prompts are too vague

Solution: Add more specific context and requirements to your input

Issue: Prompts don't match my use case

Solution: Use domain-specific templates and customize placeholders

Issue: AI responses don't meet expectations

Solution: Refine prompt with more constraints and examples

Issue: Prompts are too complex

Solution: Break into multiple simpler prompts

Advanced Features

Multi-Step Prompts

Create sequences of prompts for complex tasks:

  1. Analysis prompt
  2. Strategy prompt
  3. Implementation prompt
  4. Testing prompt
  5. Refinement prompt

Prompt Chaining

Connect multiple prompts where output of one becomes input for the next

Conditional Prompts

Create prompts that adapt based on specific conditions or previous responses

Template Variables

Use placeholders for reusable templates:

  • [TOPIC] - Main subject
  • [AUDIENCE] - Target audience
  • [LENGTH] - Desired length
  • [TONE] - Tone/style
  • [FORMAT] - Output format

Feedback & Improvement

To improve this skill:

  1. Try generated prompts with various AI tools
  2. Note which prompt structures work best
  3. Suggest new templates or categories
  4. Report any issues or ambiguities
  5. Share successful prompt examples

Notes

  • This skill works best with specific, well-defined inputs
  • More context leads to better prompts
  • Iterate and refine based on results
  • Save successful prompts as templates
  • Adapt templates to your specific needs
  • Consider the AI model's capabilities and limitations

平台分布

OpenClaw

85.95%
按下载量换算2,706

安全审计

VirusTotal

通过

ClawScan

通过

Static analysis

通过

权限和风险

需要联网

该 Skill 可能需要联网访问来源站点、仓库或外部 API;具体网络访问范围需要结合源码和 README 复核。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。目前只有一个公开来源,建议同时核对源仓库。

继续浏览同类 Skills