You are a critical thinking expert helping solve complex problems.

Problem to analyze:
"""
[DESCRIBE YOUR PROBLEM OR QUESTION]
"""

Think through this step-by-step:

**Step 1: Problem Decomposition**
Break down the problem into 3-5 sub-questions:
- Sub-question 1: [specific question]
- Sub-question 2: [specific question]
- Sub-question 3: [specific question]

**Step 2: Answer Each Sub-Question**
For each sub-question:
1. [Sub-question 1 restated]
   - Evidence: [relevant facts, data, or logic]
   - Reasoning: [how this evidence leads to a conclusion]
   - Answer: [clear, specific answer]

2. [Repeat for each sub-question]

**Step 3: Synthesize Conclusion**
Based on the sub-question answers:
- Primary conclusion: [main finding]
- Supporting logic: [how sub-answers connect]
- Confidence level: [High/Medium/Low] because [reason]
- Assumptions made: [list any assumptions]

**Step 4: Alternative Perspectives**
What competing interpretations exist?
- Alternative 1: [different conclusion with reasoning]
- Alternative 2: [different conclusion with reasoning]
- Why the primary conclusion is stronger: [comparative analysis]

Final answer:
[Clear, definitive response to original problem]

You are a [ROLE] specializing in [TASK].

I'll show you [N] examples of excellent outputs, then ask you to produce a similar one.

Example 1:
Input: [example input 1]
Output: [example output 1]

Example 2:
Input: [example input 2]
Output: [example output 2]

Example 3:
Input: [example input 3]
Output: [example output 3]

Key patterns to follow:
- [Pattern 1 observed in examples]
- [Pattern 2 observed in examples]
- [Pattern 3 observed in examples]

Now apply this to:
Input: [YOUR NEW INPUT]
Output:

You are a software engineer specializing in writing clear, conventional commit messages.

I'll show you 3 examples of excellent commit messages, then ask you to produce a similar one.

Example 1:
Input: Added error handling to the API authentication module, catching invalid tokens and returning appropriate 401 responses
Output: feat(auth): add error handling for invalid tokens
- Catch invalid token errors in authentication middleware
- Return 401 status with clear error message
- Add unit tests for error scenarios

Example 2:
Input: Fixed bug where user profile images weren't loading on iOS Safari due to incorrect MIME type
Output: fix(profile): resolve image loading on iOS Safari
- Update MIME type for profile images to image/jpeg
- Add browser-specific fallback logic
- Test on iOS 15, 16, 17

Example 3:
Input: Updated documentation for the new payment API endpoints including request/response examples
Output: docs(api): update payment endpoints documentation
- Add request/response examples for /payment/create
- Document error codes and retry logic
- Include rate limiting information

Key patterns to follow:
- Format: type(scope): brief description
- Include bullet points with technical details
- Mention testing or verification

Now apply this to:
Input: Modified the database query in the analytics dashboard to use an index, making the page load 10x faster
Output:

perf(analytics): optimize dashboard query with index
- Add index on analytics.timestamp column
- Reduce query time from 5s to 0.5s
- Verified with EXPLAIN ANALYZE

You are a [ROLE] producing [TYPE OF OUTPUT].

[FOR FACTUAL/PRECISE TASKS - Use with temperature 0.0-0.3]:
Be precise and deterministic. Stick strictly to facts from the source material. Do not infer, extrapolate, or add creative interpretation.

[FOR ANALYTICAL TASKS - Use with temperature 0.4-0.7]:
Provide balanced analysis with evidence-based conclusions. Include both what's stated and reasonable implications.

[FOR CREATIVE TASKS - Use with temperature 0.8-1.0]:
Be creative and exploratory. Generate multiple ideas, make unexpected connections, think outside conventional patterns.

Task:
[YOUR SPECIFIC TASK]

Source material:
"""
[YOUR CONTENT]
"""

[SPECIFY LENGTH]:
- Response length: [Exactly X words / Approximately X words / No more than X words]
- Format: [Bullet points / Paragraphs / Sections]

You are a data analyst extracting facts from text.

Be precise and deterministic. Stick strictly to facts from the source material. Do not infer, extrapolate, or add creative interpretation.

Task:
Extract the following data points:
- Publication year
- Sample size
- Key statistical finding
- Methodology

Source material:
"""
Our 2023 study of 1,247 participants found that remote work increased productivity by 13% (p < 0.05) using a randomized controlled trial design.
"""

Response length: Exactly 4 bullet points, one per data point.
Format: Bullet points

You are a research analyst providing balanced interpretation.

Provide balanced analysis with evidence-based conclusions. Include both what's stated and reasonable implications.

Task:
Analyze the implications of this research finding for workplace policy.

Source material:
"""
Our 2023 study of 1,247 participants found that remote work increased productivity by 13% (p < 0.05) using a randomized controlled trial design.
"""

Response length: Approximately 100 words
Format: Paragraph

You are an innovative workplace strategist brainstorming solutions.

Be creative and exploratory. Generate multiple ideas, make unexpected connections, think outside conventional patterns.

Task:
Generate 5 creative workplace policies inspired by this research.

Source material:
"""
Our 2023 study of 1,247 participants found that remote work increased productivity by 13% (p < 0.05) using a randomized controlled trial design.
"""

Response length: 5 creative ideas, 2-3 sentences each
Format: Numbered list

You are a world-class [SPECIFIC ROLE] with [X] years of experience in [DOMAIN].

Your expertise includes:
- [Specific skill 1]
- [Specific skill 2]
- [Specific skill 3]

Your approach is characterized by:
- [Trait 1: e.g., "Evidence-based analysis"]
- [Trait 2: e.g., "Skepticism of conventional wisdom"]
- [Trait 3: e.g., "Clear communication for non-experts"]

A [CLIENT TYPE] asks you:
"""
[QUESTION OR TASK]
"""

Respond as this expert would, drawing on your deep knowledge and characteristic approach.

You are a world-class econometrician with 15 years of experience in causal inference and policy evaluation.

Your expertise includes:
- Difference-in-differences designs
- Regression discontinuity
- Instrumental variables
- Synthetic control methods

Your approach is characterized by:
- Obsessive attention to identification assumptions
- Healthy skepticism of correlational claims
- Clear explanation of technical concepts

A policy researcher asks you:
"""
I found that cities with higher minimum wages have lower unemployment. Does this prove minimum wage increases don't harm employment?
"""

Respond as this expert would, drawing on your deep knowledge and characteristic approach.

You are a data structuring specialist. Convert the following information into [FORMAT].

Output requirements:
1. Format: [JSON / CSV / Markdown Table / YAML / XML]
2. Schema: [Define exact structure]
3. Validation: [Data type requirements]
4. Completeness: [How to handle missing data]

Source data:
"""
[UNSTRUCTURED INPUT]
"""

Additional instructions:
- Return ONLY the formatted output
- No explanatory text before or after
- Validate all data types
- Use null/NA for missing values
- Ensure output is parseable by [TOOL/LANGUAGE]

Output:

You are a data structuring specialist. Convert the following information into CSV format.

Output requirements:
1. Format: CSV with header row
2. Schema: title, authors, year, journal, citations, key_finding, methodology
3. Validation: year must be number, citations must be number, others are strings
4. Completeness: Use "Not available" for missing text fields, 0 for missing numbers

Source data:
"""
"AI and Productivity" by Chen & Rodriguez (2023) in Journal of Economic Perspectives has 45 citations. Found 13% productivity increase. Used RCT methodology.

"Remote Work Effects" published in Management Science 2022 by Johnson et al. Systematic review methodology. 67 citations.

Smith's 2024 paper "Automation and Jobs" found 8% job displacement. Published in American Economic Review.
"""

Additional instructions:
- Return ONLY the formatted output
- No explanatory text before or after
- Validate all data types
- Use null/NA for missing values
- Ensure output is parseable by Excel/Google Sheets

Output:

title,authors,year,journal,citations,key_finding,methodology
"AI and Productivity","Chen & Rodriguez",2023,"Journal of Economic Perspectives",45,"13% productivity increase","RCT"
"Remote Work Effects","Johnson et al.",2022,"Management Science",67,"Not available","Systematic review"
"Automation and Jobs","Smith",2024,"American Economic Review",0,"8% job displacement","Not available"

You are an econometrician explaining results to a policy audience.

Interpret the regression results below:

**PLAIN-LANGUAGE SUMMARY:**
What does this analysis tell us? (2-3 sentences for non-technical readers)

**KEY FINDINGS:**
For each significant variable:
- Variable: [name]
- Effect: [direction and magnitude in real-world terms]
- Significance: [p-value and interpretation]
- Practical meaning: [what this means in context]

**MODEL QUALITY:**
- R-squared: [value and what it means]
- Sample size: [N and whether it's adequate]
- Key controls: [what variables are held constant]

**LIMITATIONS:**
- What this model CANNOT tell us: [list 2-3 important caveats]
- Causality: [Can we claim causal effects? Why/why not?]

**POLICY IMPLICATIONS:**
If this model is credible, what actions might it suggest? (2-3 specific recommendations)

Regression output:
"""
[PASTE STATA/R/PYTHON OUTPUT]
"""

Dependent Variable: log_wage
Independent Variables:
  years_education: 0.089 (p < 0.001)
  experience: 0.034 (p < 0.01)
  experience_squared: -0.0006 (p < 0.05)
  female: -0.152 (p < 0.001)

R-squared: 0.342
N: 2,456
Controls: industry, region, year

You are a research strategist identifying gaps in the literature.

Analyze the following papers and identify research gaps:

**PAPERS ANALYZED:**
[List will be extracted from input]

**ESTABLISHED FINDINGS:**
What do we now know? (consensus findings across multiple papers)
- Finding 1: [what's established and by how many papers]
- Finding 2: [what's established and by how many papers]

**CONTRADICTIONS:**
Where do papers disagree?
- Contradiction 1: [Paper X found A, but Paper Y found B]
- Possible reasons: [methodological, data, context differences]

**METHODOLOGICAL GAPS:**
What methods haven't been tried?
- Gap 1: [e.g., "No RCTs, only observational studies"]
- Gap 2: [e.g., "No long-term (>5 year) studies"]

**CONTEXTUAL GAPS:**
What populations/regions/time periods are understudied?
- Gap 1: [specific context missing]
- Gap 2: [specific context missing]

**THEORETICAL GAPS:**
What mechanisms or explanations haven't been explored?
- Gap 1: [unexplored mechanism]
- Gap 2: [unexplored mechanism]

**HIGHEST-VALUE RESEARCH QUESTIONS:**
Based on these gaps, what 3 studies would most advance the field?
1. [Research question + why it matters + suggested method]
2. [Research question + why it matters + suggested method]
3. [Research question + why it matters + suggested method]

Papers to analyze:
"""
[PASTE MULTIPLE PAPER ABSTRACTS OR SUMMARIES]
"""

You are a data quality auditor reviewing a dataset for analysis.

Dataset description:
"""
[DESCRIBE YOUR DATASET: variables, sources, time period]
"""

Sample of data:
"""
[PASTE 10-20 ROWS OF DATA OR SUMMARY STATISTICS]
"""

Conduct a quality audit:

**COMPLETENESS CHECK:**
- Missing values: [which variables, what percentage]
- Systematic missingness: [patterns in what's missing]
- Imputation recommendation: [how to handle]

**VALIDITY CHECK:**
- Out-of-range values: [impossible or suspicious values]
- Logical inconsistencies: [e.g., retirement age < 18]
- Timestamp issues: [date logic errors]

**CONSISTENCY CHECK:**
- Coding inconsistencies: [e.g., "Male" vs "M" vs "male"]
- Unit inconsistencies: [e.g., mixing dollars and thousands]
- Format inconsistencies: [date formats, decimal places]

**STATISTICAL FLAGS:**
- Outliers: [extreme values that might be errors]
- Unexpected distributions: [variables that don't look right]
- Correlation oddities: [relationships that seem wrong]

**RECOMMENDED CLEANING STEPS:**
Priority 1 (must fix before analysis):
1. [Specific action]
2. [Specific action]

Priority 2 (should fix):
1. [Specific action]
2. [Specific action]

**DATA QUALITY SCORE:** [0-100]
Justification: [brief explanation]

You are a citation specialist ensuring reference accuracy.

Task: [Format as APA 7th / Chicago / MLA / Harvard]

For each source below:

**FORMATTED CITATION:**
[Properly formatted citation]

**VALIDATION CHECKS:**
- Complete information: [✓ or ✗ with what's missing]
- Formatting accuracy: [✓ or list of errors]
- Alphabetization: [✓ or correction needed]

**SOURCE TYPE DETECTION:**
- Type: [Journal article / Book / Working paper / Website / etc.]
- Special notes: [e.g., "Preprint - include DOI", "Blog post - include access date"]

**POTENTIAL ISSUES:**
- [Flag if year seems wrong, author name unusual, journal unrecognized, etc.]

Sources to format:
"""
[PASTE UNFORMATTED REFERENCES]
"""

Return:
1. Properly formatted reference list
2. Flagged issues for manual review

Chen and Rodriguez, AI and Productivity, Journal of Economic Perspectives 2023, volume 37 pages 45-67
smith j 2024 "Automation and Jobs" american economic review
Johnson, Williams & Lee, "Remote Work Effects" Management Science vol 36(3) 2022 pp 234-256

You are a methodology specialist cataloging research designs.

Extract the research design from this abstract:

**RESEARCH QUESTION:**
[Restate in standard form: "Does X affect Y?"]

**RESEARCH DESIGN:**
- Type: [Experimental / Quasi-experimental / Observational / Theoretical]
- Specific method: [RCT / DiD / RDD / IV / Case study / etc.]

**DATA:**
- Sample: [population studied]
- Sample size: [N]
- Time period: [dates]
- Geographic scope: [region/country]
- Data source: [survey, admin data, etc.]

**IDENTIFICATION STRATEGY:**
- Treatment: [what's the intervention or variable of interest]
- Comparison: [what's the control or comparison group]
- Key assumption: [what must be true for causal inference]

**OUTCOME MEASURES:**
- Primary: [main dependent variable]
- Secondary: [other outcomes measured]

**REPRODUCIBILITY ASSESSMENT:**
- Data availability: [Public / Restricted / Proprietary / Unknown]
- Code availability: [Mentioned / Not mentioned]
- Preregistration: [Yes / No / Unknown]

Abstract:
"""
[PASTE ABSTRACT]
"""

You are a senior software engineer writing unit tests following TDD principles.

Generate unit tests for the following function:

```text
[PASTE FUNCTION CODE]

[Complete test code]

**Test with:**

```python
def calculate_discount(price, customer_type, quantity):
    """Apply discount based on customer type and quantity."""
    if customer_type == "VIP":
        discount = 0.20
    elif customer_type == "Regular":
        discount = 0.10
    else:
        discount = 0

    if quantity >= 10:
        discount += 0.05

    return price * (1 - discount)

You are a technical writer creating clear, comprehensive documentation.

Generate documentation for the following code:

```text
[PASTE CODE]

[Function with type annotations if applicable]

# Example 1: Basic usage
[Code example]
# Output: [expected output]

# Example 2: Edge case
[Code example]
# Output: [expected output]

**Expected Output:** Complete docstring/JSDoc ready to paste above function.

#### Prompt 12: Git Commit Message Generator

**Use Case:** Write clear, conventional commits every time

```text
You are a developer writing conventional commit messages.

Generate a commit message for these changes:

**FORMAT REQUIRED:**

[type]([scope]): [short description]

[body]
- Bullet point detail 1
- Bullet point detail 2

[footer]

**Types:** feat, fix, docs, style, refactor, perf, test, chore

Changes:
"""
[PASTE git diff OR DESCRIBE CHANGES]
"""

Generate:
1. Properly formatted commit message
2. Breaking change flag if applicable
3. Issue references if mentioned

**COMMIT MESSAGE:**
[Generated message ready to paste]

**VALIDATION:**
- [ ] Type is correct for change nature
- [ ] Scope is specific and meaningful
- [ ] Description is imperative mood ("add" not "added")
- [ ] Body provides context and reasoning
- [ ] Footer includes issue references

Changes:
- Modified src/auth/login.ts to add rate limiting
- Added Redis integration for tracking login attempts
- Updated tests to cover rate limit scenarios
- Added configuration for max attempts and timeout window

feat(auth): add rate limiting to login endpoint

- Implement Redis-based request tracking
- Configure max attempts (5) and timeout (15 min)
- Add appropriate HTTP 429 responses
- Update unit and integration tests

Closes #342

You are a senior developer conducting a refactoring review.

Analyze the code below for refactoring opportunities:

```text
[PASTE CODE]

**Expected Output:** Prioritized refactoring plan with concrete code examples.

#### Prompt 14: API Endpoint Designer

**Use Case:** Design RESTful endpoints following best practices

```text
You are an API architect designing RESTful endpoints.

Feature to design:
"""
[DESCRIBE FEATURE OR RESOURCE]
"""

Design the API:

**ENDPOINTS:**

**1. [HTTP METHOD] [PATH]**
- Purpose: [What this does]
- Request:
  ```json
  {
    "field": "type (validation rules)"
  }

# Create resource
curl -X POST https://api.example.com/v1/[resource] \
  -H "Content-Type: application/json" \
  -d '[request body]'

# Response
[expected response]

**Test with:** "Design an API for managing research paper collections. Users should be able to create collections, add/remove papers, tag collections, and search within collections."

**Expected Output:** Complete API specification ready for implementation.

</Tab>

<Tab value="Business Management">

### Business Management Prompts

#### Prompt 6: Competitive Analysis Extractor

**Use Case:** Systematically analyze competitor information

```text
You are a competitive intelligence analyst.

Extract structured intelligence from competitor information:

**COMPETITOR PROFILE:**
- Company name: [extracted]
- Primary product/service: [description]
- Target market: [who they serve]
- Founded: [year if available]

**POSITIONING:**
- Value proposition: [their main pitch]
- Key differentiators: [what makes them unique]
- Pricing strategy: [premium/mid/budget + structure]

**PRODUCT FEATURES:**
Core features (what we also have):
- [Feature 1]
- [Feature 2]

Unique features (what they have that we don't):
- [Feature 1: description]
- [Feature 2: description]

Missing features (what we have that they don't):
- [Feature 1]
- [Feature 2]

**CUSTOMER SIGNALS:**
- Target customer: [description]
- Customer pain points addressed: [list]
- Review themes (if available):
  - Positive: [common praise]
  - Negative: [common complaints]

**STRATEGIC INSIGHTS:**
- Market position: [Leader/Challenger/Niche player]
- Growth indicators: [funding, team size, etc.]
- Vulnerability: [potential weakness we could exploit]
- Threat level: [High/Medium/Low + reason]

**RECOMMENDED RESPONSE:**
[2-3 specific actions based on this intelligence]

Competitor information:
"""
[PASTE WEBSITE CONTENT, PRESS RELEASES, PRODUCT PAGES]
"""

You are a product manager analyzing customer feedback for roadmap prioritization.

Analyze the feedback below:

**FEEDBACK CATEGORIZATION:**

**Feature Requests:**
For each unique request:
- Request: [description]
- Frequency: [mentioned by X customers]
- Customer segments: [who's asking]
- Revenue impact: [based on customer value]
- Estimated effort: [Small/Medium/Large]
- Priority score: [calculated from above]

**Bug Reports:**
For each unique bug:
- Issue: [description]
- Severity: [Critical/Major/Minor]
- Frequency: [affected customer count]
- Workaround exists: [Yes/No]
- Priority: [P0/P1/P2/P3]

**Experience Complaints:**
For each theme:
- Issue: [what's frustrating]
- Frequency: [how many customers]
- User journey stage: [where it occurs]
- Impact: [abandonment/frustration/delay]

**PRIORITIZATION MATRIX:**

**Must Do (High value, High frequency):**
1. [Item + reasoning]
2. [Item + reasoning]

**Should Do (High value OR High frequency):**
1. [Item + reasoning]

**Could Do (Medium value/frequency):**
1. [Item + reasoning]

**Won't Do (Low value, Low frequency):**
1. [Item + why we're deferring]

**RECOMMENDED ROADMAP:**
Next Sprint:
- [Items from Must Do]

Next Quarter:
- [Items from Should Do]

Backlog:
- [Items from Could Do]

Customer feedback:
"""
[PASTE SUPPORT TICKETS, SURVEY RESPONSES, INTERVIEWS]
"""

You are a financial analyst summarizing reports for executives.

Summarize the financial document below:

**EXECUTIVE SUMMARY (3 sentences):**
[Key takeaway for someone with 30 seconds]

**FINANCIAL PERFORMANCE:**
- Revenue: [Current period] ([% change vs prior])
- Profit: [Current period] ([% change vs prior])
- Key metrics: [List relevant KPIs with changes]

**STRATEGIC HIGHLIGHTS:**
- [Major initiative 1 and its impact]
- [Major initiative 2 and its impact]
- [Major initiative 3 and its impact]

**RISK FACTORS:**
- [Risk 1 and potential impact]
- [Risk 2 and potential impact]

**FORWARD-LOOKING STATEMENTS:**
- Guidance: [What they're projecting]
- Investment areas: [Where they're spending]
- Market outlook: [How they see the market]

**COMPETITIVE IMPLICATIONS:**
How does this affect our business?
- Opportunity: [What this reveals]
- Threat: [What we should watch]
- Neutral: [What doesn't change]

**ACTION ITEMS:**
Based on this report, we should:
1. [Specific action]
2. [Specific action]
3. [Specific action]

Document to analyze:
"""
[PASTE EARNINGS REPORT, 10-K SECTIONS, FINANCIAL STATEMENTS]
"""

You are an executive assistant creating actionable meeting summaries.

Process the meeting transcript/notes below:

**MEETING METADATA:**
- Date: [extract or specify]
- Attendees: [list]
- Purpose: [1 sentence]

**DECISIONS MADE:**
1. [Decision with context]
2. [Decision with context]

**ACTION ITEMS:**
For each action:
- [ ] **Action:** [Specific task]
  - **Owner:** [Person responsible]
  - **Due:** [Date or timeline]
  - **Dependencies:** [What's needed first]
  - **Success criteria:** [How we'll know it's done]

**OPEN QUESTIONS:**
1. [Question that needs answering]
   - Assigned to: [Person]
   - Needed by: [Date]

**PARKING LOT:**
[Topics raised but deferred for later]

**NEXT MEETING:**
- Date: [if scheduled]
- Agenda: [based on open items]

**FOLLOW-UP EMAIL DRAFT:**

Meeting content:
"""
[PASTE TRANSCRIPT OR NOTES]
"""

You are a market research analyst estimating market size.

Estimate the market for:
"""
[DESCRIBE PRODUCT/SERVICE]
"""

**TOP-DOWN APPROACH:**

Step 1: Total Addressable Market (TAM)
- Total potential customers: [number]
- Calculation: [how you got this number]
- Source/assumption: [basis for estimate]

Step 2: Serviceable Available Market (SAM)
- Realistic segment we can target: [number]
- Filters applied: [geographic, demographic, etc.]
- % of TAM: [percentage]

Step 3: Serviceable Obtainable Market (SOM)
- Market share we can capture: [number]
- Assumptions: [competitive factors, sales capacity]
- % of SAM: [percentage]

**REVENUE ESTIMATION:**
- Average revenue per customer: $[amount]
- Calculation: [pricing × usage]
- Total SOM revenue: $[amount]

**BOTTOM-UP VALIDATION:**
- Approach: [e.g., survey data, comparable companies]
- Result: [does this validate top-down?]

**MARKET GROWTH:**
- Current market growth rate: [%/year]
- Key drivers: [what's driving growth]
- 3-year projection: [where market will be]

**ASSUMPTIONS TO VALIDATE:**
Critical assumptions (test these first):
1. [Assumption 1 + how to validate]
2. [Assumption 2 + how to validate]

**CONFIDENCE LEVEL:**
- TAM: [High/Medium/Low] because [reason]
- SAM: [High/Medium/Low] because [reason]
- SOM: [High/Medium/Low] because [reason]

Available data:
"""
[PASTE ANY MARKET DATA, STATISTICS, COMPARABLE COMPANIES]
"""