@ckelsoe/prompt-architect 3.0.2 → 3.1.1

package/skills/prompt-architect/references/frameworks/chain-of-thought.md

@@ -1,424 +1,424 @@
-# Chain of Thought (CoT) Framework
-
-## Overview
-
-Chain of Thought is a prompting technique that encourages step-by-step reasoning and makes the thinking process explicit. Instead of jumping to answers, it guides Claude to break down complex problems, show intermediate steps, and verify logic along the way.
-
-## Core Concept
-
-CoT works by:
-1. Breaking complex problems into smaller steps
-2. Making reasoning explicit at each step
-3. Showing work rather than just final answers
-4. Verifying logic progressively
-5. Enabling error detection and correction
-
-## When to Use Chain of Thought
-
-**Ideal for:**
-- Mathematical problem-solving
-- Logical reasoning tasks
-- Complex multi-step analysis
-- Debugging and troubleshooting
-- Decision-making with trade-offs
-- Tasks where "showing your work" matters
-
-**Not needed for:**
-- Simple lookups or facts
-- Straightforward transformations
-- Tasks with obvious single steps
-- Creative writing without logic requirements
-
-## Implementation Approaches
-
-### Approach 1: Explicit Instruction
-Directly instruct Claude to think step-by-step.
-
-**Example:**
-```
-Solve this problem step-by-step, showing your reasoning at each stage:
-
-A company's revenue grew 15% in Q1, decreased 8% in Q2, grew 20% in Q3, and grew 5% in Q4. If they started the year with $1M in quarterly revenue, what was their Q4 revenue?
-
-Think through this carefully:
-1. Calculate Q1 revenue
-2. Calculate Q2 revenue
-3. Calculate Q3 revenue
-4. Calculate Q4 revenue
-5. Verify your calculation
-```
-
-### Approach 2: Zero-Shot CoT
-Use trigger phrases that activate reasoning.
-
-**Magic phrases:**
-- "Let's think step by step."
-- "Let's work through this systematically."
-- "Let's break this down."
-- "Think carefully about this."
-
-**Example:**
-```
-What's the best database choice for a real-time analytics platform handling 1M events/second with complex aggregations? Let's think step by step.
-```
-
-### Approach 3: Few-Shot CoT
-Provide examples showing step-by-step reasoning.
-
-**Example:**
-```
-Example 1:
-Q: If a train travels 120 miles in 2 hours, what's its speed?
-A: Let's break this down:
-1. Speed = Distance / Time
-2. Distance = 120 miles
-3. Time = 2 hours
-4. Speed = 120 / 2 = 60 mph
-Therefore: 60 mph
-
-Example 2:
-Q: A store offers 20% off, then an additional 10% off. What's the total discount on a $100 item?
-A: Let's work through this:
-1. First discount: 20% of $100 = $20
-2. Price after first discount: $100 - $20 = $80
-3. Second discount: 10% of $80 = $8
-4. Final price: $80 - $8 = $72
-5. Total discount: $100 - $72 = $28
-6. Total discount percentage: 28/100 = 28%
-Therefore: 28% total discount (not 30%)
-
-Now solve:
-Q: [Your problem]
-A: Let's break this down:
-```
-
-### Approach 4: Structured CoT
-Provide a template for reasoning.
-
-**Example:**
-```
-Analyze whether we should migrate to microservices using this reasoning structure:
-
-STEP 1 - CURRENT STATE:
-[Describe current architecture]
-
-STEP 2 - PROBLEMS:
-[What problems exist?]
-
-STEP 3 - MICROSERVICES BENEFITS:
-[How would microservices help?]
-
-STEP 4 - MICROSERVICES COSTS:
-[What are the downsides?]
-
-STEP 5 - ALTERNATIVES:
-[What else could we do?]
-
-STEP 6 - COMPARISON:
-[Compare options systematically]
-
-STEP 7 - RECOMMENDATION:
-[Final decision with reasoning]
-```
-
-## Best Practices
-
-### 1. Be Explicit About Steps
-```
-Good:
-"Calculate the ROI step by step:
-1. Calculate total costs
-2. Calculate total revenue
-3. Calculate profit (revenue - costs)
-4. Calculate ROI (profit / costs * 100)
-5. Interpret the result"
-
-Poor:
-"Calculate the ROI"
-```
-
-### 2. Request Verification
-```
-Good:
-"After solving, verify your answer makes sense and check your math."
-
-Poor:
-[No verification step]
-```
-
-### 3. Encourage Self-Correction
-```
-Good:
-"If you find an error in your reasoning, acknowledge it and correct your approach."
-
-Poor:
-[No mention of error handling]
-```
-
-### 4. Ask for Uncertainty Acknowledgment
-```
-Good:
-"If you're uncertain at any step, state your assumptions and why you made them."
-
-Poor:
-[Assumes certainty at all steps]
-```
-
-## Complete Examples
-
-### Example 1: Debugging
-
-**Without CoT:**
-```
-Why isn't this code working?
-[code snippet]
-```
-
-**With CoT:**
-```
-Debug this code by thinking through it step-by-step:
-
-[code snippet]
-
-Use this process:
-1. Read the code and state what it's supposed to do
-2. Identify the input and expected output
-3. Trace through execution line by line
-4. Note any suspicious patterns or red flags
-5. Form hypotheses about the bug
-6. Test each hypothesis against the code
-7. Identify the root cause
-8. Propose a fix
-9. Verify the fix addresses the issue
-```
-
-### Example 2: System Design
-
-**Without CoT:**
-```
-Design a URL shortener.
-```
-
-**With CoT:**
-```
-Design a URL shortener by working through these steps:
-
-STEP 1 - REQUIREMENTS:
-Think through what the system needs to do.
-- Functional requirements?
-- Non-functional requirements?
-- Scale expectations?
-
-STEP 2 - API DESIGN:
-What endpoints do we need?
-- Consider: creation, retrieval, analytics
-
-STEP 3 - DATA MODEL:
-How should we store URLs?
-- What fields are needed?
-- What indexes?
-
-STEP 4 - URL GENERATION:
-How do we create short codes?
-- Base62 encoding? Hash? Counter?
-- Collision handling?
-
-STEP 5 - SCALABILITY:
-How does this scale?
-- Bottlenecks?
-- Caching strategy?
-- Database sharding?
-
-STEP 6 - TRADE-OFFS:
-What are we optimizing for vs. sacrificing?
-
-For each step, explain your reasoning.
-```
-
-### Example 3: Decision Making
-
-**Without CoT:**
-```
-Should we use MongoDB or PostgreSQL?
-```
-
-**With CoT:**
-```
-Decide between MongoDB and PostgreSQL by reasoning through:
-
-STEP 1 - USE CASE ANALYSIS:
-What are we building and what data patterns will we have?
-- Data structure (structured/unstructured?)
-- Relationships (complex/simple?)
-- Query patterns (known/unknown?)
-
-STEP 2 - MONGODB EVALUATION:
-Pros for our use case:
-[List and explain]
-Cons for our use case:
-[List and explain]
-
-STEP 3 - POSTGRESQL EVALUATION:
-Pros for our use case:
-[List and explain]
-Cons for our use case:
-[List and explain]
-
-STEP 4 - CRITICAL REQUIREMENTS:
-What are our non-negotiables?
-- [Requirement 1 and why it matters]
-- [Requirement 2 and why it matters]
-
-STEP 5 - SCORING:
-Rate each database against our requirements.
-
-STEP 6 - DECISION:
-Based on the analysis above, choose one and explain why.
-
-STEP 7 - VALIDATE:
-Does this decision make sense? Any red flags?
-```
-
-## Combining CoT with Other Frameworks
-
-### CoT + CO-STAR
-```
-CONTEXT: [Situation]
-OBJECTIVE: [Goal]
-...
-
-PROCESS (Chain of Thought):
-Work through this step-by-step:
-1. [Step 1]
-2. [Step 2]
-...
-```
-
-### CoT + RISEN
-```
-ROLE: [Expertise]
-INSTRUCTIONS: Use step-by-step reasoning for each stage
-STEPS:
-1. [Step 1] - Think through: [reasoning points]
-2. [Step 2] - Consider: [reasoning points]
-...
-```
-
-### CoT + RISE
-```
-ROLE: [Analyst]
-INPUT: [Data]
-STEPS:
-1. [Action] - Reasoning: [Why this step?]
-2. [Action] - Reasoning: [What are we looking for?]
-...
-```
-
-## Advanced Techniques
-
-### Self-Consistency
-Generate multiple reasoning paths and choose the most common answer:
-
-```
-Solve this problem using three different approaches:
-
-APPROACH 1:
-[Method 1 with step-by-step reasoning]
-
-APPROACH 2:
-[Method 2 with step-by-step reasoning]
-
-APPROACH 3:
-[Method 3 with step-by-step reasoning]
-
-COMPARISON:
-Which approach is most reliable and why?
-
-FINAL ANSWER:
-Based on the most consistent result.
-```
-
-### Uncertainty Quantification
-```
-For each step, rate your confidence:
-1. [Step 1] - Confidence: High/Medium/Low - Because: [reason]
-2. [Step 2] - Confidence: High/Medium/Low - Because: [reason]
-
-If any step has low confidence, explore alternatives.
-```
-
-### Backward Chaining
-```
-Start with the desired outcome and work backwards:
-
-GOAL: [What we want to achieve]
-
-STEP -1: To achieve this, what must be true immediately before?
-STEP -2: To achieve that, what must be true before it?
-STEP -3: Keep working backwards...
-STEP -N: What's our starting point?
-
-Now verify the forward path makes sense.
-```
-
-## Common Patterns
-
-### Mathematical Problem Solving
-```
-1. Understand: Restate the problem in your own words
-2. Identify: What do we know? What do we need to find?
-3. Plan: What formula or approach applies?
-4. Execute: Work through the math step-by-step
-5. Verify: Does the answer make sense? Check units, magnitude
-```
-
-### Logical Reasoning
-```
-1. Premises: State what we know to be true
-2. Inference: What can we deduce from premises?
-3. Test: Do our deductions hold up?
-4. Conclusion: What can we conclude?
-5. Verify: Are there any logical fallacies?
-```
-
-### Troubleshooting
-```
-1. Observe: What's the symptom?
-2. Hypothesize: What could cause this?
-3. Test: How can we test each hypothesis?
-4. Eliminate: Rule out what doesn't fit
-5. Identify: What's the root cause?
-6. Solve: How do we fix it?
-7. Verify: Did it work?
-```
-
-## Quality Indicators
-
-**Good CoT reasoning shows:**
-- ✅ Each step builds on previous ones
-- ✅ Assumptions are stated explicitly
-- ✅ Calculations/logic are shown
-- ✅ Intermediate results are verified
-- ✅ Uncertainties are acknowledged
-- ✅ Final answer follows logically
-
-**Poor CoT reasoning shows:**
-- ❌ Jumping to conclusions
-- ❌ Skipping intermediate steps
-- ❌ Hiding assumptions
-- ❌ No verification
-- ❌ Unjustified leaps in logic
-
-## Assessment Checklist
-
-When using Chain of Thought:
-- [ ] Problem is complex enough to warrant CoT
-- [ ] Steps are explicitly requested
-- [ ] Reasoning structure is provided or suggested
-- [ ] Verification is included
-- [ ] Self-correction is encouraged
-- [ ] Each step is actionable
-- [ ] Logic flow is clear
-- [ ] Assumptions are acknowledged
-- [ ] Final answer connects to reasoning
-- [ ] Work shown, not just answer
+# Chain of Thought (CoT) Framework
+
+## Overview
+
+Chain of Thought is a prompting technique that encourages step-by-step reasoning and makes the thinking process explicit. Instead of jumping to answers, it guides Claude to break down complex problems, show intermediate steps, and verify logic along the way.
+
+## Core Concept
+
+CoT works by:
+1. Breaking complex problems into smaller steps
+2. Making reasoning explicit at each step
+3. Showing work rather than just final answers
+4. Verifying logic progressively
+5. Enabling error detection and correction
+
+## When to Use Chain of Thought
+
+**Ideal for:**
+- Mathematical problem-solving
+- Logical reasoning tasks
+- Complex multi-step analysis
+- Debugging and troubleshooting
+- Decision-making with trade-offs
+- Tasks where "showing your work" matters
+
+**Not needed for:**
+- Simple lookups or facts
+- Straightforward transformations
+- Tasks with obvious single steps
+- Creative writing without logic requirements
+
+## Implementation Approaches
+
+### Approach 1: Explicit Instruction
+Directly instruct Claude to think step-by-step.
+
+**Example:**
+```
+Solve this problem step-by-step, showing your reasoning at each stage:
+
+A company's revenue grew 15% in Q1, decreased 8% in Q2, grew 20% in Q3, and grew 5% in Q4. If they started the year with $1M in quarterly revenue, what was their Q4 revenue?
+
+Think through this carefully:
+1. Calculate Q1 revenue
+2. Calculate Q2 revenue
+3. Calculate Q3 revenue
+4. Calculate Q4 revenue
+5. Verify your calculation
+```
+
+### Approach 2: Zero-Shot CoT
+Use trigger phrases that activate reasoning.
+
+**Magic phrases:**
+- "Let's think step by step."
+- "Let's work through this systematically."
+- "Let's break this down."
+- "Think carefully about this."
+
+**Example:**
+```
+What's the best database choice for a real-time analytics platform handling 1M events/second with complex aggregations? Let's think step by step.
+```
+
+### Approach 3: Few-Shot CoT
+Provide examples showing step-by-step reasoning.
+
+**Example:**
+```
+Example 1:
+Q: If a train travels 120 miles in 2 hours, what's its speed?
+A: Let's break this down:
+1. Speed = Distance / Time
+2. Distance = 120 miles
+3. Time = 2 hours
+4. Speed = 120 / 2 = 60 mph
+Therefore: 60 mph
+
+Example 2:
+Q: A store offers 20% off, then an additional 10% off. What's the total discount on a $100 item?
+A: Let's work through this:
+1. First discount: 20% of $100 = $20
+2. Price after first discount: $100 - $20 = $80
+3. Second discount: 10% of $80 = $8
+4. Final price: $80 - $8 = $72
+5. Total discount: $100 - $72 = $28
+6. Total discount percentage: 28/100 = 28%
+Therefore: 28% total discount (not 30%)
+
+Now solve:
+Q: [Your problem]
+A: Let's break this down:
+```
+
+### Approach 4: Structured CoT
+Provide a template for reasoning.
+
+**Example:**
+```
+Analyze whether we should migrate to microservices using this reasoning structure:
+
+STEP 1 - CURRENT STATE:
+[Describe current architecture]
+
+STEP 2 - PROBLEMS:
+[What problems exist?]
+
+STEP 3 - MICROSERVICES BENEFITS:
+[How would microservices help?]
+
+STEP 4 - MICROSERVICES COSTS:
+[What are the downsides?]
+
+STEP 5 - ALTERNATIVES:
+[What else could we do?]
+
+STEP 6 - COMPARISON:
+[Compare options systematically]
+
+STEP 7 - RECOMMENDATION:
+[Final decision with reasoning]
+```
+
+## Best Practices
+
+### 1. Be Explicit About Steps
+```
+Good:
+"Calculate the ROI step by step:
+1. Calculate total costs
+2. Calculate total revenue
+3. Calculate profit (revenue - costs)
+4. Calculate ROI (profit / costs * 100)
+5. Interpret the result"
+
+Poor:
+"Calculate the ROI"
+```
+
+### 2. Request Verification
+```
+Good:
+"After solving, verify your answer makes sense and check your math."
+
+Poor:
+[No verification step]
+```
+
+### 3. Encourage Self-Correction
+```
+Good:
+"If you find an error in your reasoning, acknowledge it and correct your approach."
+
+Poor:
+[No mention of error handling]
+```
+
+### 4. Ask for Uncertainty Acknowledgment
+```
+Good:
+"If you're uncertain at any step, state your assumptions and why you made them."
+
+Poor:
+[Assumes certainty at all steps]
+```
+
+## Complete Examples
+
+### Example 1: Debugging
+
+**Without CoT:**
+```
+Why isn't this code working?
+[code snippet]
+```
+
+**With CoT:**
+```
+Debug this code by thinking through it step-by-step:
+
+[code snippet]
+
+Use this process:
+1. Read the code and state what it's supposed to do
+2. Identify the input and expected output
+3. Trace through execution line by line
+4. Note any suspicious patterns or red flags
+5. Form hypotheses about the bug
+6. Test each hypothesis against the code
+7. Identify the root cause
+8. Propose a fix
+9. Verify the fix addresses the issue
+```
+
+### Example 2: System Design
+
+**Without CoT:**
+```
+Design a URL shortener.
+```
+
+**With CoT:**
+```
+Design a URL shortener by working through these steps:
+
+STEP 1 - REQUIREMENTS:
+Think through what the system needs to do.
+- Functional requirements?
+- Non-functional requirements?
+- Scale expectations?
+
+STEP 2 - API DESIGN:
+What endpoints do we need?
+- Consider: creation, retrieval, analytics
+
+STEP 3 - DATA MODEL:
+How should we store URLs?
+- What fields are needed?
+- What indexes?
+
+STEP 4 - URL GENERATION:
+How do we create short codes?
+- Base62 encoding? Hash? Counter?
+- Collision handling?
+
+STEP 5 - SCALABILITY:
+How does this scale?
+- Bottlenecks?
+- Caching strategy?
+- Database sharding?
+
+STEP 6 - TRADE-OFFS:
+What are we optimizing for vs. sacrificing?
+
+For each step, explain your reasoning.
+```
+
+### Example 3: Decision Making
+
+**Without CoT:**
+```
+Should we use MongoDB or PostgreSQL?
+```
+
+**With CoT:**
+```
+Decide between MongoDB and PostgreSQL by reasoning through:
+
+STEP 1 - USE CASE ANALYSIS:
+What are we building and what data patterns will we have?
+- Data structure (structured/unstructured?)
+- Relationships (complex/simple?)
+- Query patterns (known/unknown?)
+
+STEP 2 - MONGODB EVALUATION:
+Pros for our use case:
+[List and explain]
+Cons for our use case:
+[List and explain]
+
+STEP 3 - POSTGRESQL EVALUATION:
+Pros for our use case:
+[List and explain]
+Cons for our use case:
+[List and explain]
+
+STEP 4 - CRITICAL REQUIREMENTS:
+What are our non-negotiables?
+- [Requirement 1 and why it matters]
+- [Requirement 2 and why it matters]
+
+STEP 5 - SCORING:
+Rate each database against our requirements.
+
+STEP 6 - DECISION:
+Based on the analysis above, choose one and explain why.
+
+STEP 7 - VALIDATE:
+Does this decision make sense? Any red flags?
+```
+
+## Combining CoT with Other Frameworks
+
+### CoT + CO-STAR
+```
+CONTEXT: [Situation]
+OBJECTIVE: [Goal]
+...
+
+PROCESS (Chain of Thought):
+Work through this step-by-step:
+1. [Step 1]
+2. [Step 2]
+...
+```
+
+### CoT + RISEN
+```
+ROLE: [Expertise]
+INSTRUCTIONS: Use step-by-step reasoning for each stage
+STEPS:
+1. [Step 1] - Think through: [reasoning points]
+2. [Step 2] - Consider: [reasoning points]
+...
+```
+
+### CoT + RISE
+```
+ROLE: [Analyst]
+INPUT: [Data]
+STEPS:
+1. [Action] - Reasoning: [Why this step?]
+2. [Action] - Reasoning: [What are we looking for?]
+...
+```
+
+## Advanced Techniques
+
+### Self-Consistency
+Generate multiple reasoning paths and choose the most common answer:
+
+```
+Solve this problem using three different approaches:
+
+APPROACH 1:
+[Method 1 with step-by-step reasoning]
+
+APPROACH 2:
+[Method 2 with step-by-step reasoning]
+
+APPROACH 3:
+[Method 3 with step-by-step reasoning]
+
+COMPARISON:
+Which approach is most reliable and why?
+
+FINAL ANSWER:
+Based on the most consistent result.
+```
+
+### Uncertainty Quantification
+```
+For each step, rate your confidence:
+1. [Step 1] - Confidence: High/Medium/Low - Because: [reason]
+2. [Step 2] - Confidence: High/Medium/Low - Because: [reason]
+
+If any step has low confidence, explore alternatives.
+```
+
+### Backward Chaining
+```
+Start with the desired outcome and work backwards:
+
+GOAL: [What we want to achieve]
+
+STEP -1: To achieve this, what must be true immediately before?
+STEP -2: To achieve that, what must be true before it?
+STEP -3: Keep working backwards...
+STEP -N: What's our starting point?
+
+Now verify the forward path makes sense.
+```
+
+## Common Patterns
+
+### Mathematical Problem Solving
+```
+1. Understand: Restate the problem in your own words
+2. Identify: What do we know? What do we need to find?
+3. Plan: What formula or approach applies?
+4. Execute: Work through the math step-by-step
+5. Verify: Does the answer make sense? Check units, magnitude
+```
+
+### Logical Reasoning
+```
+1. Premises: State what we know to be true
+2. Inference: What can we deduce from premises?
+3. Test: Do our deductions hold up?
+4. Conclusion: What can we conclude?
+5. Verify: Are there any logical fallacies?
+```
+
+### Troubleshooting
+```
+1. Observe: What's the symptom?
+2. Hypothesize: What could cause this?
+3. Test: How can we test each hypothesis?
+4. Eliminate: Rule out what doesn't fit
+5. Identify: What's the root cause?
+6. Solve: How do we fix it?
+7. Verify: Did it work?
+```
+
+## Quality Indicators
+
+**Good CoT reasoning shows:**
+- ✅ Each step builds on previous ones
+- ✅ Assumptions are stated explicitly
+- ✅ Calculations/logic are shown
+- ✅ Intermediate results are verified
+- ✅ Uncertainties are acknowledged
+- ✅ Final answer follows logically
+
+**Poor CoT reasoning shows:**
+- ❌ Jumping to conclusions
+- ❌ Skipping intermediate steps
+- ❌ Hiding assumptions
+- ❌ No verification
+- ❌ Unjustified leaps in logic
+
+## Assessment Checklist
+
+When using Chain of Thought:
+- [ ] Problem is complex enough to warrant CoT
+- [ ] Steps are explicitly requested
+- [ ] Reasoning structure is provided or suggested
+- [ ] Verification is included
+- [ ] Self-correction is encouraged
+- [ ] Each step is actionable
+- [ ] Logic flow is clear
+- [ ] Assumptions are acknowledged
+- [ ] Final answer connects to reasoning
+- [ ] Work shown, not just answer