ralph-prompt-research

Compare original and translation side by side

🇺🇸

Original

English
🇨🇳

Translation

Chinese

Ralph Prompt Generator: Research & Analysis

Ralph提示词生成器:研究与分析

Overview

概述

Generates structured prompts for research, analysis, and planning tasks using the Ralph Wiggum technique. These prompts define clear research methodology, required sources, synthesis requirements, and concrete deliverables rather than code.
Best For:
  • Codebase analysis and audits
  • Security vulnerability assessments
  • Technology research and comparison
  • Migration planning
  • Architecture design documents
  • Performance analysis
  • Dependency audits
  • Documentation generation from code
Ralph Philosophy: This generator embraces the principle that failures are deterministic and fixable. Each research iteration builds on previous findings. Don't fear incomplete first passes—they're expected and provide data for deeper analysis in subsequent iterations.
使用Ralph Wiggum技术为研究、分析与规划任务生成结构化提示词。这些提示词定义了清晰的研究方法、所需来源、综合要求以及具体交付物,而非代码。
适用场景:
  • 代码库分析与审计
  • 安全漏洞评估
  • 技术调研与对比
  • 迁移计划制定
  • 架构设计文档编写
  • 性能分析
  • 依赖审计
  • 基于代码生成文档
Ralph核心理念:该生成器秉持“失败是可预测且可修复的”原则。每一次研究迭代都基于之前的发现展开。无需担心首次输出不完整——这是预期之内的,且能为后续更深层次的分析提供数据支持。

Quick Start

快速开始

Input Required:
  1. Research objective (what you need to understand/analyze)
  2. Scope (what to include/exclude)
  3. Deliverable format (report, plan, document)
  4. Completion promise
Generate prompt with:
Generate a Ralph research prompt for:
Objective: [What to research/analyze]
Scope: [What's included]
Deliverable: [What to produce]
Promise: [COMPLETION_PHRASE]
所需输入:
  1. 研究目标(需要理解/分析的内容)
  2. 范围(包含/排除的内容)
  3. 交付物格式(报告、计划、文档)
  4. 完成标识
生成提示词的命令:
Generate a Ralph research prompt for:
Objective: [What to research/analyze]
Scope: [What's included]
Deliverable: [What to produce]
Promise: [COMPLETION_PHRASE]

Prompt Generation Workflow

提示词生成流程

Step 1: Define Research Scope

步骤1:定义研究范围

Research Definition Template:
markdown
OBJECTIVE: [What question to answer or what to analyze]
SCOPE:
  Include: [What's in scope]
  Exclude: [What's explicitly out of scope]
SOURCES: [Where to look for information]
DELIVERABLE: [What artifact to produce]
SUCCESS: [How to know research is complete]
研究定义模板:
markdown
OBJECTIVE: [需要解答的问题或分析的内容]
SCOPE:
  Include: [包含的内容]
  Exclude: [明确排除的内容]
SOURCES: [信息来源]
DELIVERABLE: [需要产出的成果]
SUCCESS: [判断研究完成的标准]

Step 2: Map Research Phases

步骤2:规划研究阶段

Standard research phases:
PhaseNamePurposeDeliverable
1DiscoveryIdentify what existsInventory/map
2AnalysisDeep dive into findingsAnalysis notes
3SynthesisDraw conclusionsFindings summary
4RecommendationsActionable next stepsRecommendation doc
5DocumentationFinal deliverableComplete report
标准研究阶段:
阶段名称目标交付物
1发现明确现有内容清单/映射图
2分析深入研究发现的内容分析笔记
3综合得出结论发现总结
4建议提供可执行的下一步方案建议文档
5文档化最终交付物完整报告

Step 3: Define Completion Criteria

步骤3:定义完成标准

Research completion criteria must be:
  • Based on coverage (examined all relevant items)
  • Based on depth (sufficient analysis per item)
  • Based on deliverable completeness (all sections written)
Examples:
Research TypeCompletion Criteria
Security auditAll OWASP categories checked, all findings documented
Codebase analysisAll modules examined, architecture documented
Tech comparisonAll options evaluated against all criteria
Migration planAll systems inventoried, migration steps defined
研究完成标准必须满足:
  • 基于覆盖范围(检查所有相关内容)
  • 基于分析深度(对每个内容进行充分分析)
  • 基于交付物完整性(所有章节已完成)
示例:
研究类型完成标准
安全审计覆盖所有OWASP分类,所有发现已记录
代码库分析所有模块已检查,架构已记录
技术对比所有选项已按所有标准评估
迁移计划所有系统已盘点,迁移步骤已定义

Step 4: Structure the Prompt

步骤4:构建提示词结构

Use this template:
markdown
undefined
使用以下模板:
markdown
undefined

Research: [Research Title]

研究:[研究标题]

Objective

目标

[Clear statement of what this research aims to answer or analyze]
[清晰说明本次研究旨在解答的问题或分析的内容]

Scope

范围

In Scope

包含内容

  • [Item 1 to analyze]
  • [Item 2 to analyze]
  • [Item 3 to analyze]
  • [待分析内容1]
  • [待分析内容2]
  • [待分析内容3]

Out of Scope

排除内容

  • [Explicitly excluded 1]
  • [Explicitly excluded 2]
  • [明确排除内容1]
  • [明确排除内容2]

Sources to Examine

参考来源

  • [Source type 1]: [Where to find]
  • [Source type 2]: [Where to find]
  • [Source type 3]: [Where to find]

  • [来源类型1]:[获取位置]
  • [来源类型2]:[获取位置]
  • [来源类型3]:[获取位置]

Phase 1: Discovery

阶段1:发现

Objective

目标

Create comprehensive inventory of [what to inventory].
创建[待盘点内容]的全面清单。

Tasks

任务

  1. [Discovery task 1]
  2. [Discovery task 2]
  3. [Discovery task 3]
  1. [发现任务1]
  2. [发现任务2]
  3. [发现任务3]

Deliverable: Inventory

交付物:清单

Create
[filename].md
with:
markdown
undefined
创建
[filename].md
,格式如下:
markdown
undefined

[Inventory Title]

[清单标题]

Items Found

已发现内容

#ItemLocationTypeNotes
1
2
undefined
序号内容位置类型备注
1
2
undefined

Phase 1 Success Criteria

阶段1完成标准

  • All [scope items] examined
  • Inventory file created
  • No areas unexplored
  • 所有[范围内内容]已检查
  • 清单文件已创建
  • 无未探索领域

Phase 1 Checkpoint

阶段1检查点

Document in inventory file:
DISCOVERY COMPLETE:
- Items found: [count]
- Areas examined: [list]
- Coverage: Complete/Partial

在清单文件中记录:
发现阶段完成:
- 已发现内容数量:[count]
- 已检查领域:[列表]
- 覆盖范围:完整/部分

Phase 2: Analysis

阶段2:分析

Objective

目标

Deep analysis of each discovered item.
对每个发现的内容进行深度分析。

Analysis Framework

分析框架

For each item, analyze:
  • [Dimension 1]: [What to examine]
  • [Dimension 2]: [What to examine]
  • [Dimension 3]: [What to examine]
对每个内容,需分析:
  • [维度1]:[检查内容]
  • [维度2]:[检查内容]
  • [维度3]:[检查内容]

Tasks

任务

  1. Analyze each item against framework
  2. Document findings per item
  3. Identify patterns across items
  4. Note anomalies or concerns
  1. 按照框架分析每个内容
  2. 记录每个内容的分析结果
  3. 识别内容间的模式
  4. 记录异常或问题

Deliverable: Analysis Notes

交付物:分析笔记

Create
[analysis-filename].md
with:
markdown
undefined
创建
[analysis-filename].md
,格式如下:
markdown
undefined

Analysis: [Item Name]

分析:[内容名称]

[Dimension 1]

[维度1]

[Findings]
[分析结果]

[Dimension 2]

[维度2]

[Findings]
[分析结果]

[Dimension 3]

[维度3]

[Findings]
[分析结果]

Summary

总结

[Key takeaways for this item]
undefined
[该内容的关键要点]
undefined

Phase 2 Success Criteria

阶段2完成标准

  • All inventory items analyzed
  • Analysis notes for each item
  • Patterns identified
  • Anomalies documented
  • 所有清单内容已分析
  • 每个内容都有分析笔记
  • 已识别模式
  • 异常已记录

Phase 2 Checkpoint

阶段2检查点

ANALYSIS COMPLETE:
- Items analyzed: [X/Y]
- Key patterns: [list]
- Concerns identified: [count]

分析阶段完成:
- 已分析内容数量:[X/Y]
- 关键模式:[列表]
- 已识别问题数量:[count]

Phase 3: Synthesis

阶段3:综合

Objective

目标

Synthesize findings into coherent conclusions.
将分析结果整合为连贯的结论。

Tasks

任务

  1. Review all analysis notes
  2. Identify overarching themes
  3. Draw conclusions from patterns
  4. Prioritize findings by impact
  5. Formulate key insights
  1. 回顾所有分析笔记
  2. 识别核心主题
  3. 从模式中得出结论
  4. 按影响优先级排序发现结果
  5. 提炼关键见解

Deliverable: Findings Summary

交付物:发现总结

Create
[findings-filename].md
with:
markdown
undefined
创建
[findings-filename].md
,格式如下:
markdown
undefined

Research Findings: [Topic]

研究发现:[主题]

Executive Summary

执行摘要

[2-3 paragraphs summarizing key findings]
[2-3段总结关键发现]

Key Findings

主要发现

  1. [Finding 1]: [Description and evidence]
  2. [Finding 2]: [Description and evidence]
  3. [Finding 3]: [Description and evidence]
  1. [发现1]:[描述及证据]
  2. [发现2]:[描述及证据]
  3. [发现3]:[描述及证据]

Patterns Observed

观察到的模式

  • [Pattern 1]
  • [Pattern 2]
  • [模式1]
  • [模式2]

Areas of Concern

问题领域

  1. [问题1]:[详情]
  2. [问题2]:[详情]

Data Summary

数据总结

  • Total items examined: [X]
  • Categories: [list]
  • Analysis coverage: [X%]
undefined
  • 已检查内容总数:[X]
  • 分类:[列表]
  • 分析覆盖范围:[X%]
undefined

Phase 3 Success Criteria

阶段3完成标准

  • All findings synthesized
  • Conclusions supported by evidence
  • Priorities established
  • Summary document complete

  • 所有发现已整合
  • 结论有证据支持
  • 已确定优先级
  • 总结文档已完成

Phase 4: Recommendations

阶段4:建议

Objective

目标

Provide actionable recommendations based on findings.
基于发现结果提供可执行的建议。

Tasks

任务

  1. Develop recommendations for each finding
  2. Prioritize by impact and effort
  3. Identify dependencies between recommendations
  4. Create action timeline (if applicable)
  1. 为每个发现结果制定建议
  2. 按影响和实施难度排序
  3. 识别建议间的依赖关系
  4. 创建实施时间线(如适用)

Deliverable: Recommendations

交付物:建议

Add to findings document or create separate:
markdown
undefined
添加至发现总结文档或单独创建:
markdown
undefined

Recommendations

建议

High Priority (Address Immediately)

高优先级(立即处理)

  1. [Recommendation]
    • Rationale: [Why]
    • Effort: [Low/Medium/High]
    • Impact: [Low/Medium/High]
    • Dependencies: [Any]
  1. [建议内容]
    • 理由:[原因]
    • 实施难度:[低/中/高]
    • 影响:[低/中/高]
    • 依赖关系:[如有]

Medium Priority (Plan for Soon)

中优先级(近期规划)

[...]
[...]

Low Priority (Future Consideration)

低优先级(未来考虑)

[...]
[...]

Implementation Order

实施顺序

  1. First: [Recommendation X]
  2. Then: [Recommendation Y]
  3. Finally: [Recommendation Z]
undefined
  1. 首先:[建议X]
  2. 其次:[建议Y]
  3. 最后:[建议Z]
undefined

Phase 4 Success Criteria

阶段4完成标准

  • All findings have recommendations
  • Priorities assigned
  • Implementation order defined
  • Dependencies identified

  • 所有发现结果都有对应建议
  • 已分配优先级
  • 已定义实施顺序
  • 已识别依赖关系

Phase 5: Final Documentation

阶段5:最终文档

Objective

目标

Produce final polished deliverable.
产出最终的完善交付物。

Tasks

任务

  1. Compile all sections into final document
  2. Add executive summary
  3. Review for completeness
  4. Ensure actionability
  5. Format consistently
  1. 将所有章节整合为最终文档
  2. 添加执行摘要
  3. 检查完整性
  4. 确保内容可执行
  5. 统一格式

Deliverable: Final Report

交付物:最终报告

Create
[final-report].md
with all sections:
  1. Executive Summary
  2. Scope and Methodology
  3. Findings
  4. Analysis
  5. Recommendations
  6. Appendix (raw data if needed)
创建
[final-report].md
,包含以下所有章节:
  1. 执行摘要
  2. 范围与方法
  3. 发现结果
  4. 分析内容
  5. 建议
  6. 附录(如需包含原始数据)

Phase 5 Success Criteria

阶段5完成标准

  • All sections complete
  • Executive summary written
  • Consistent formatting
  • No TODO items
  • Actionable and clear

  • 所有章节已完成
  • 执行摘要已撰写
  • 格式统一
  • 无待办项
  • 内容清晰且可执行

Final Verification

最终验证

Completeness Checklist

完整性检查清单

  • Phase 1: Discovery complete, inventory exists
  • Phase 2: All items analyzed, notes exist
  • Phase 3: Findings synthesized, summary exists
  • Phase 4: Recommendations provided, prioritized
  • Phase 5: Final document complete
  • 阶段1:发现完成,清单已存在
  • 阶段2:所有内容已分析,笔记已存在
  • 阶段3:发现结果已整合,总结已存在
  • 阶段4:已提供建议并排序
  • 阶段5:最终文档已完成

Quality Checklist

质量检查清单

  • All conclusions supported by evidence
  • Recommendations are actionable
  • Document is well-organized
  • No sections are missing content
  • Executive summary captures key points
  • 所有结论有证据支持
  • 建议可执行
  • 文档结构清晰
  • 无内容缺失章节
  • 执行摘要涵盖关键要点

Completion

完成标识

When ALL of the following are true:
  • All 5 phases completed
  • All deliverables created
  • Completeness checklist verified
  • Quality checklist verified
Output: <promise>[COMPLETION_PHRASE]</promise>
当满足以下所有条件时:
  • 所有5个阶段已完成
  • 所有交付物已创建
  • 完整性检查清单已验证
  • 质量检查清单已验证
输出:<promise>[COMPLETION_PHRASE]</promise>

Progress Tracking

进度跟踪

RESEARCH PROGRESS:
Phase 1 (Discovery): [COMPLETE/IN PROGRESS]
Phase 2 (Analysis): [COMPLETE/IN PROGRESS/NOT STARTED]
Phase 3 (Synthesis): [COMPLETE/IN PROGRESS/NOT STARTED]
Phase 4 (Recommendations): [COMPLETE/IN PROGRESS/NOT STARTED]
Phase 5 (Documentation): [COMPLETE/IN PROGRESS/NOT STARTED]

Items examined: [X/Y]
Deliverables created: [list]
研究进度:
阶段1(发现):[已完成/进行中]
阶段2(分析):[已完成/进行中/未开始]
阶段3(综合):[已完成/进行中/未开始]
阶段4(建议):[已完成/进行中/未开始]
阶段5(文档化):[已完成/进行中/未开始]

已检查内容数量:[X/Y]
已创建交付物:[列表]

If Stuck

遇到瓶颈时

  • Document what information is missing
  • List sources already examined
  • Identify the specific gap
  • Do NOT skip analysis or output false promise
undefined
  • 记录缺失的信息
  • 列出已检查的来源
  • 明确具体的信息缺口
  • 请勿跳过分析或输出虚假完成标识
undefined

Example Prompts

示例提示词

Example 1: Security Audit

示例1:安全审计

Input:
  • Objective: Identify security vulnerabilities in authentication module
  • Scope: src/auth/ directory
  • Deliverable: Security audit report with findings and fixes
  • Promise: SECURITY_AUDIT_COMPLETE
Generated Prompt:
markdown
undefined
输入:
  • 目标:识别认证模块中的安全漏洞
  • 范围:src/auth/目录
  • 交付物:包含发现结果与修复方案的安全审计报告
  • 完成标识:SECURITY_AUDIT_COMPLETE
生成的提示词:
markdown
undefined

Research: Authentication Security Audit

研究:认证模块安全审计

Objective

目标

Identify security vulnerabilities in the authentication module following OWASP guidelines and provide remediation recommendations.
遵循OWASP指南识别认证模块中的安全漏洞,并提供修复建议。

Scope

范围

In Scope

包含内容

  • All files in src/auth/
  • Authentication flows (login, signup, password reset)
  • Session management
  • Token handling
  • Input validation
  • src/auth/下的所有文件
  • 认证流程(登录、注册、密码重置)
  • 会话管理
  • Token处理
  • 输入验证

Out of Scope

排除内容

  • Authorization (permissions, roles)
  • Third-party auth providers
  • Frontend security
  • 授权(权限、角色)
  • 第三方认证提供商
  • 前端安全

Sources to Examine

参考来源

  • Code files: src/auth/**/*.ts
  • Tests: tests/auth/**/*.ts
  • Configuration: auth-related config files
  • Dependencies: auth-related npm packages

  • 代码文件:src/auth/**/*.ts
  • 测试用例:tests/auth/**/*.ts
  • 配置文件:认证相关配置文件
  • 依赖包:认证相关npm包

Phase 1: Discovery

阶段1:发现

Tasks

任务

  1. List all files in auth module
  2. Identify all entry points (routes)
  3. Map authentication flows
  4. Identify external dependencies
  5. List all sensitive operations
  1. 列出认证模块中的所有文件
  2. 识别所有入口点(路由)
  3. 绘制认证流程图
  4. 识别外部依赖
  5. 列出所有敏感操作

Deliverable: auth-inventory.md

交付物:auth-inventory.md

markdown
undefined
markdown
undefined

Auth Module Inventory

认证模块清单

Files

文件

FilePurposeEntry PointsSensitive Ops
文件用途入口点敏感操作

Dependencies

依赖包

| Package | Version | Security Notes |
| 包名 | 版本 | 安全说明 |

Authentication Flows

认证流程

  1. Login: [path]
  2. Signup: [path] [...]
undefined
  1. 登录:[路径]
  2. 注册:[路径] [...]
undefined

Phase 1 Success

阶段1完成标准

  • All auth files listed
  • All entry points identified
  • Flows mapped
  • Dependencies inventoried

  • 所有认证文件已列出
  • 所有入口点已识别
  • 流程已绘制
  • 依赖包已盘点

Phase 2: Analysis

阶段2:分析

Security Analysis Framework

安全分析框架

For each component, check:
  • A1 Injection: SQL, NoSQL, command injection risks
  • A2 Broken Auth: Weak passwords, session issues
  • A3 Sensitive Data: Password storage, token exposure
  • A4 XXE: XML parsing vulnerabilities
  • A5 Access Control: Proper authorization checks
  • A6 Misconfiguration: Insecure defaults
  • A7 XSS: User input handling
  • A8 Deserialization: Unsafe object handling
  • A9 Components: Vulnerable dependencies
  • A10 Logging: Insufficient monitoring
对每个组件,检查以下内容:
  • A1 注入:SQL、NoSQL、命令注入风险
  • A2 身份认证漏洞:弱密码、会话问题
  • A3 敏感数据暴露:密码存储、Token泄露
  • A4 XXE攻击:XML解析漏洞
  • A5 访问控制:适当的授权检查
  • A6 配置错误:不安全的默认配置
  • A7 XSS攻击:用户输入处理
  • A8 反序列化:不安全的对象处理
  • A9 组件漏洞:存在漏洞的依赖包
  • A10 日志不足:监控不充分

Tasks

任务

  1. Analyze each file against OWASP framework
  2. Check password hashing implementation
  3. Review JWT token handling
  4. Examine session management
  5. Check input validation
  1. 按照OWASP框架分析每个文件
  2. 检查密码哈希实现
  3. 审查JWT Token处理
  4. 检查会话管理
  5. 检查输入验证

Deliverable: security-analysis.md

交付物:security-analysis.md

Per-file analysis with vulnerabilities found.

每个文件的分析结果及已发现的漏洞。

Phase 3: Synthesis

阶段3:综合

Tasks

任务

  1. Categorize all vulnerabilities by severity
  2. Identify systemic issues
  3. Calculate risk scores
  4. Prioritize by exploitability
  1. 按严重程度对所有漏洞分类
  2. 识别系统性问题
  3. 计算风险评分
  4. 按可利用性排序

Deliverable: security-findings.md

交付物:security-findings.md

markdown
undefined
markdown
undefined

Security Findings

安全发现结果

Critical (Requires Immediate Action)

严重(需立即处理)

  • [Vulnerability]: [Location] - [Impact]
  • [漏洞]:[位置] - [影响]

High

高风险

[...]
[...]

Medium

中风险

[...]
[...]

Low

低风险

[...]
[...]

Summary Statistics

统计摘要

  • Critical: X
  • High: X
  • Medium: X
  • Low: X

---
  • 严重:X
  • 高风险:X
  • 中风险:X
  • 低风险:X

---

Phase 4: Recommendations

阶段4:建议

Tasks

任务

  1. Provide fix for each vulnerability
  2. Prioritize fixes by risk
  3. Estimate effort for each fix
  4. Identify quick wins
  1. 为每个漏洞提供修复方案
  2. 按风险优先级排序修复方案
  3. 估算每个修复方案的实施难度
  4. 识别快速优化项

Deliverable: remediation-plan.md

交付物:remediation-plan.md

markdown
undefined
markdown
undefined

Remediation Plan

修复计划

Immediate Actions

立即执行

  1. [Fix]: [How] - [Effort]
  1. [修复方案]:[实施方式] - [难度]

Short-term (This Sprint)

短期(当前迭代)

[...]
[...]

Medium-term

中期

[...]

---
[...]

---

Phase 5: Final Report

阶段5:最终报告

Deliverable: AUTH-SECURITY-AUDIT.md

交付物:AUTH-SECURITY-AUDIT.md

Complete report with:
  1. Executive Summary
  2. Methodology
  3. Findings by Severity
  4. Detailed Vulnerability Descriptions
  5. Remediation Plan
  6. Verification Steps

完整报告包含:
  1. 执行摘要
  2. 研究方法
  3. 按严重程度分类的发现结果
  4. 详细漏洞描述
  5. 修复计划
  6. 验证步骤

Completion

完成标识

When audit report complete with all vulnerabilities documented and remediation provided:
Output: <promise>SECURITY_AUDIT_COMPLETE</promise>
undefined
当审计报告完成,所有漏洞已记录且提供修复方案时:
输出:<promise>SECURITY_AUDIT_COMPLETE</promise>
undefined

Example 2: Technology Comparison

示例2:技术框架对比

Input:
  • Objective: Compare React vs Vue vs Svelte for new frontend
  • Scope: Team skills, project requirements, ecosystem
  • Deliverable: Comparison report with recommendation
  • Promise: FRAMEWORK_COMPARISON_COMPLETE
Generated Prompt:
markdown
undefined
输入:
  • 目标:为新前端项目对比React、Vue与Svelte
  • 范围:团队技能、项目需求、生态系统
  • 交付物:包含建议的对比报告
  • 完成标识:FRAMEWORK_COMPARISON_COMPLETE
生成的提示词:
markdown
undefined

Research: Frontend Framework Comparison

研究:前端框架对比

Objective

目标

Compare React, Vue, and Svelte to recommend the best framework for the new frontend project based on team skills, project requirements, and ecosystem.
对比React、Vue与Svelte,基于团队技能、项目需求与生态系统,为新前端项目推荐最优框架。

Scope

范围

In Scope

包含内容

  • Learning curve for team
  • Performance characteristics
  • Ecosystem (libraries, tools)
  • Community and support
  • Project requirements fit
  • Long-term maintainability
  • 团队学习曲线
  • 性能特性
  • 生态系统(库、工具)
  • 社区与支持
  • 项目需求适配性
  • 长期可维护性

Out of Scope

排除内容

  • Mobile development (React Native, etc.)
  • SSR frameworks (Next.js, Nuxt, SvelteKit)
  • Backend considerations

  • 移动端开发(React Native等)
  • SSR框架(Next.js、Nuxt、SvelteKit)
  • 后端相关考量

Phase 1: Discovery

阶段1:发现

Tasks

任务

  1. Document team's current skills
  2. List project requirements
  3. Identify key evaluation criteria
  4. Gather framework documentation
  1. 记录团队当前技能
  2. 列出项目需求
  3. 确定关键评估标准
  4. 收集框架文档

Deliverable: evaluation-criteria.md

交付物:evaluation-criteria.md

markdown
undefined
markdown
undefined

Evaluation Framework

评估框架

Team Context

团队背景

  • Current experience: [frameworks]
  • Available learning time: [duration]
  • Team size: [N]
  • 当前经验:[掌握的框架]
  • 可用学习时间:[时长]
  • 团队规模:[N]

Project Requirements

项目需求

  • [Requirement 1]
  • [Requirement 2] [...]
  • [需求1]
  • [需求2] [...]

Evaluation Criteria

评估标准

CriterionWeightReason
Learning curve25%Team has limited React experience
Performance20%App will be data-heavy
[...]

---
标准权重理由
学习曲线25%团队仅有有限的React经验
性能20%应用将处理大量数据
[...]

---

Phase 2: Analysis

阶段2:分析

Analysis Framework

分析框架

For each framework:
  • Learning Curve: Time to productivity
  • Performance: Bundle size, runtime speed, memory
  • Ecosystem: Available libraries, tools, IDE support
  • Community: Documentation, Stack Overflow, GitHub activity
  • Hiring: Developer availability
  • Architecture: How it fits project patterns
对每个框架,分析以下内容:
  • 学习曲线:达到高效开发的时间
  • 性能:包体积、运行速度、内存占用
  • 生态系统:可用库、工具、IDE支持
  • 社区:文档、Stack Overflow、GitHub活跃度
  • 招聘难度:开发者可获得性
  • 架构适配:与项目模式的契合度

Deliverable: framework-analysis.md

交付物:framework-analysis.md

markdown
undefined
markdown
undefined

React Analysis

React分析

Learning Curve

学习曲线

[Analysis]
[分析内容]

Performance

性能

[Analysis]
[... repeat for Vue and Svelte]

---
[分析内容]
[... 重复Vue与Svelte的分析]

---

Phase 3: Synthesis

阶段3:综合

Tasks

任务

  1. Score each framework on each criterion
  2. Apply weights
  3. Calculate overall scores
  4. Identify dealbreakers
  1. 为每个框架在各标准下评分
  2. 应用权重计算
  3. 得出总分
  4. 识别否决项

Deliverable: comparison-matrix.md

交付物:comparison-matrix.md

markdown
undefined
markdown
undefined

Framework Comparison Matrix

框架对比矩阵

CriterionWeightReactVueSvelte
Learning Curve25%344
Performance20%345
[...]
Weighted TotalX.XX.XX.X

---
标准权重ReactVueSvelte
学习曲线25%344
性能20%345
[...]
加权总分X.XX.XX.X

---

Phase 4: Recommendations

阶段4:建议

Deliverable: recommendation.md

交付物:recommendation.md

markdown
undefined
markdown
undefined

Framework Recommendation

框架推荐

Recommendation: [Framework]

推荐框架:[框架名称]

Rationale

理由

[Why this framework is best for this team and project]
[该框架为何最适合当前团队与项目]

Advantages

优势

  • [Advantage 1]
  • [Advantage 2]
  • [优势1]
  • [优势2]

Concerns and Mitigations

顾虑与缓解方案

  • [Concern 1]: [How to address]
  • [顾虑1]:[解决方式]

Implementation Approach

实施步骤

  1. [Step 1]
  2. [Step 2] [...]

---
  1. [步骤1]
  2. [步骤2] [...]

---

Phase 5: Final Report

阶段5:最终报告

Deliverable: FRAMEWORK-COMPARISON-REPORT.md

交付物:FRAMEWORK-COMPARISON-REPORT.md



Completion

完成标识

When complete comparison with recommendation:
Output: <promise>FRAMEWORK_COMPARISON_COMPLETE</promise>
undefined
当包含推荐的完整对比报告完成时:
输出:<promise>FRAMEWORK_COMPARISON_COMPLETE</promise>
undefined

Best Practices

最佳实践

Research Methodology

研究方法

  • Define scope explicitly (what's in/out)
  • Use consistent analysis framework
  • Document evidence for conclusions
  • Separate findings from opinions
  • 明确定义范围边界(包含/排除内容)
  • 使用统一的分析框架
  • 为结论记录证据
  • 将发现结果与观点分开

Deliverables

交付物

  • Create artifacts as you go
  • Use consistent format
  • Include data to support conclusions
  • Make recommendations actionable
  • 边研究边创建成果
  • 使用统一格式
  • 包含支持结论的数据
  • 确保建议可执行

DO:

注意事项:

  • Define clear scope boundaries
  • Create deliverables at each phase
  • Support conclusions with evidence
  • Provide actionable recommendations
  • Track progress systematically
  • 明确定义范围边界
  • 在每个阶段创建交付物
  • 用证据支持结论
  • 提供可执行的建议
  • 系统跟踪进度

DON'T:

禁止事项:

  • Skip the discovery phase
  • Make recommendations without evidence
  • Leave sections incomplete
  • Output promise before documentation
  • 跳过发现阶段
  • 在无证据支持的情况下给出建议
  • 保留未完成章节
  • 在文档完成前输出完成标识

Integration with Ralph Loop

与Ralph Loop集成

bash
/ralph-wiggum:ralph-loop "[paste generated prompt]" --completion-promise "YOUR_PROMISE" --max-iterations 50
Recommended iterations by research complexity:
  • Simple audit (single module):
    --max-iterations 30-40
  • Comparison (3-5 options):
    --max-iterations 40-60
  • Full codebase analysis:
    --max-iterations 70-100
  • Complex planning:
    --max-iterations 60-80

For single-task prompts, see
ralph-prompt-single-task
. For multi-task prompts, see
ralph-prompt-multi-task
. For project-level prompts, see
ralph-prompt-project
.
bash
/ralph-wiggum:ralph-loop "[粘贴生成的提示词]" --completion-promise "YOUR_PROMISE" --max-iterations 50
按研究复杂度推荐迭代次数:
  • 简单审计(单个模块):
    --max-iterations 30-40
  • 技术对比(3-5个选项):
    --max-iterations 40-60
  • 完整代码库分析:
    --max-iterations 70-100
  • 复杂规划:
    --max-iterations 60-80

如需单任务提示词,请查看
ralph-prompt-single-task
。 如需多任务提示词,请查看
ralph-prompt-multi-task
。 如需项目级提示词,请查看
ralph-prompt-project