ThoughtMCP

# Quality Checking **Time**: 10 minutes **Difficulty**: Beginner **Tools Used**: `analyze_reasoning` ## The Scenario You want to improve the quality of reasoning and decision-making by identifying biases, logical flaws, and areas for improvement. This is like having a reasoning coach that helps you think better. ## Basic Reasoning Analysis ### Analyzing Your Own Reasoning **Scenario**: You've made a decision and want to check if your reasoning was sound. **Your Reasoning Steps:** ```json { "tool": "analyze_reasoning", "arguments": { "reasoning_steps": [ { "type": "problem_identification", "content": "Our website is loading slowly, affecting user experience", "confidence": 0.9 }, { "type": "cause_analysis", "content": "The problem is probably the database queries taking too long", "confidence": 0.6 }, { "type": "solution_proposal", "content": "We should add database indexes to speed up queries", "confidence": 0.8 } ] } } ``` **What the Analysis Reveals:** ```json { "analysis": { "logical_coherence": 0.7, "evidence_support": 0.5, "bias_indicators": [ { "type": "confirmation_bias", "severity": "moderate", "explanation": "Jumped to database conclusion without investigating other causes" } ], "reasoning_quality": "moderate", "improvement_suggestions": [ "Gather performance metrics before assuming root cause", "Consider alternative explanations (network, frontend, CDN)", "Test hypothesis before implementing solution" ] } } ``` ### Analyzing Complex Decisions **Scenario**: You're choosing between two job offers and want to check your reasoning. **Tool Call:** ```json { "tool": "analyze_reasoning", "arguments": { "reasoning_steps": [ { "type": "option_identification", "content": "I have two job offers: Company A offers $120k with remote work, Company B offers $100k but it's a startup with equity", "confidence": 1.0 }, { "type": "criteria_weighting", "content": "Salary is most important to me right now because I have student loans", "confidence": 0.8 }, { "type": "risk_assessment", "content": "Company A is safer because it's established, startups are risky", "confidence": 0.7 }, { "type": "decision", "content": "I should choose Company A because of higher salary and lower risk", "confidence": 0.8 } ] } } ``` **Analysis Results:** ```json { "analysis": { "logical_coherence": 0.8, "evidence_support": 0.6, "bias_indicators": [ { "type": "loss_aversion", "severity": "mild", "explanation": "Overweighting risk of startup failure" }, { "type": "anchoring_bias", "severity": "moderate", "explanation": "Anchored on salary numbers without considering total compensation" } ], "missing_considerations": [ "Long-term career growth potential", "Learning opportunities at startup vs established company", "Actual equity value calculation", "Work-life balance differences", "Team and culture fit" ], "improvement_suggestions": [ "Calculate total compensation including equity scenarios", "Consider 3-5 year career trajectory for each option", "Evaluate non-monetary factors like learning and growth", "Talk to current employees at both companies" ] } } ``` ## Identifying Common Biases ### Confirmation Bias **Example Reasoning:** ```json { "reasoning_steps": [ { "type": "hypothesis", "content": "I think our users prefer the blue button design", "confidence": 0.7 }, { "type": "evidence_gathering", "content": "I asked 3 colleagues and they all agreed blue looks better", "confidence": 0.8 }, { "type": "conclusion", "content": "Blue button is definitely the right choice", "confidence": 0.9 } ] } ``` **Analysis Identifies:** - **Confirmation bias**: Only sought confirming evidence - **Small sample size**: 3 colleagues isn't representative - **Selection bias**: Colleagues may have similar preferences - **Missing A/B testing**: No actual user data ### Anchoring Bias **Example Reasoning:** ```json { "reasoning_steps": [ { "type": "initial_estimate", "content": "The client mentioned their budget is around $50k for this project", "confidence": 0.8 }, { "type": "scope_analysis", "content": "Looking at the requirements, this seems like a $50k project", "confidence": 0.7 }, { "type": "pricing_decision", "content": "I'll quote $48k to be competitive", "confidence": 0.8 } ] } ``` **Analysis Identifies:** - **Anchoring bias**: Influenced by client's initial number - **Insufficient analysis**: Didn't independently estimate effort - **Undervaluing work**: May be pricing below actual value ### Availability Heuristic **Example Reasoning:** ```json { "reasoning_steps": [ { "type": "risk_assessment", "content": "Cloud services are unreliable because AWS had that big outage last month", "confidence": 0.8 }, { "type": "solution_preference", "content": "We should host everything on our own servers for reliability", "confidence": 0.7 } ] } ``` **Analysis Identifies:** - **Availability heuristic**: Recent dramatic event skewing perception - **Incomplete data**: Not considering overall reliability statistics - **False comparison**: Own servers may be less reliable than cloud ## Improving Reasoning Quality ### Before Making Important Decisions **Step 1: Document Your Reasoning** ```json { "tool": "analyze_reasoning", "arguments": { "reasoning_steps": [ // Your reasoning steps here ] } } ``` **Step 2: Review the Analysis** Look for: - Identified biases - Missing considerations - Weak evidence - Logical gaps **Step 3: Improve Your Reasoning** Address the identified issues: - Gather missing evidence - Consider alternative perspectives - Challenge your assumptions - Seek disconfirming evidence ### Iterative Improvement **Round 1: Initial Reasoning** ```json { "reasoning_steps": [ { "type": "problem_analysis", "content": "Our app crashes frequently, users are complaining", "confidence": 0.9 }, { "type": "solution", "content": "We need to hire more developers to fix bugs faster", "confidence": 0.7 } ] } ``` **Analysis Result**: "Jumping to solution without root cause analysis" **Round 2: Improved Reasoning** ```json { "reasoning_steps": [ { "type": "problem_analysis", "content": "Our app crashes frequently, users are complaining", "confidence": 0.9 }, { "type": "data_gathering", "content": "Crash logs show 80% of crashes are from one specific feature", "confidence": 0.9 }, { "type": "root_cause_analysis", "content": "The feature has a memory leak that wasn't caught in testing", "confidence": 0.8 }, { "type": "solution_evaluation", "content": "Fix the memory leak and improve testing process to catch similar issues", "confidence": 0.8 } ] } ``` **Analysis Result**: "Much improved - data-driven approach with root cause analysis" ## Advanced Analysis Features ### Confidence Calibration **Tool Call:** ```json { "tool": "analyze_reasoning", "arguments": { "reasoning_steps": [ { "type": "prediction", "content": "This marketing campaign will increase sales by 25%", "confidence": 0.9 } ], "context": { "domain": "marketing", "check_confidence_calibration": true } } } ``` **Analysis Checks:** - Is 90% confidence justified? - What evidence supports this confidence level? - Are you overconfident based on past predictions? ### Alternative Perspective Analysis **Tool Call:** ```json { "tool": "analyze_reasoning", "arguments": { "reasoning_steps": [ // Your reasoning steps ], "context": { "generate_alternatives": true, "perspective_count": 3 } } } ``` **Analysis Provides:** - Alternative interpretations of the evidence - Different solution approaches - Contrarian viewpoints to consider ## Try It Yourself ### Experiment 1: Bias Detection Analyze a recent decision you made: 1. **Document your reasoning steps** 2. **Run the analysis** 3. **Identify any biases** 4. **Consider how you could have reasoned better** ### Experiment 2: Iterative Improvement Take a current problem you're facing: 1. **Write your initial reasoning** 2. **Analyze it for flaws** 3. **Improve based on feedback** 4. **Analyze again to see improvement** ### Experiment 3: Confidence Calibration For your next prediction: 1. **Make the prediction with confidence level** 2. **Analyze if confidence is justified** 3. **Track actual outcome** 4. **Learn about your calibration patterns** ## Common Reasoning Patterns to Analyze ### Technical Decisions ```json { "reasoning_steps": [ { "type": "technology_selection", "content": "We should use React because it's popular", "confidence": 0.6 } ] } ``` **Common Issues:** - Popularity ≠ suitability - Missing requirements analysis - Not considering alternatives ### Business Decisions ```json { "reasoning_steps": [ { "type": "market_analysis", "content": "Our competitors are all raising prices, so we should too", "confidence": 0.7 } ] } ``` **Common Issues:** - Following competitors blindly - Not considering customer impact - Missing differentiation opportunities ### Personal Decisions ```json { "reasoning_steps": [ { "type": "career_choice", "content": "I should take this job because it pays more", "confidence": 0.8 } ] } ``` **Common Issues:** - Single-factor optimization - Ignoring long-term consequences - Not considering personal values ## Key Takeaways ### What Quality Analysis Provides - **Bias identification**: Spot thinking errors - **Gap detection**: Find missing considerations - **Evidence evaluation**: Assess reasoning strength - **Improvement suggestions**: Specific ways to think better ### When to Use Analysis - **Important decisions**: High-stakes choices - **Complex problems**: Multi-faceted issues - **Learning moments**: After successes or failures - **Team decisions**: Before group consensus ### Building Better Reasoning Habits 1. **Document reasoning steps** for important decisions 2. **Regularly analyze** your thinking patterns 3. **Seek disconfirming evidence** actively 4. **Consider alternative perspectives** systematically 5. **Track prediction accuracy** to improve calibration ## Next Steps - **[Complete Workflow](complete-workflow.md)** - Use all tools together - **[Real-World Examples](../real-world/)** - See complex applications - **[Integration Examples](../integration/)** - Implementation patterns --- _Ready to see all tools working together? Check out the [Complete Workflow](complete-workflow.md)._