3 AI Prompts That Save Developers 10+ Hours/Week

// Paste this into ChatGPT, Claude, Gemini, or Copilot You are a principal software engineer conducting a comprehensive code review. Analyze the following code with the same rigor you'd apply to a production system handling millions of requests. **Code to review:** {code} **Context:** - Language/framework: {language} - This code is part of: {component_description} - Known constraints: {constraints} Review the code across ALL of the following dimensions. For each issue found, specify the exact line(s), severity (Critical / Major / Minor / Nit), and provide a concrete fix — not just a description of the problem. ### 1. Correctness - Logic errors, off-by-one errors, null/undefined handling - Edge cases that would cause incorrect results - Assumptions that may not hold in production ### 2. Readability & Maintainability - Naming clarity (variables, functions, classes) - Function length and single-responsibility adherence - Dead code, unnecessary complexity, magic numbers - Missing or misleading comments ### 3. Error Handling - Unhandled exceptions or error states - Error messages that lack context for debugging - Silent failures that swallow errors - Missing retry logic for transient failures ### 4. Security - Input validation gaps - Injection vulnerabilities (SQL, XSS, command) - Hardcoded secrets or credentials - Insecure defaults or permissions ### 5. Performance - Unnecessary allocations or copies - N+1 query patterns or unbounded loops - Missing caching opportunities - Inefficient data structures for the use case ### 6. Testability - Code that is difficult to unit test - Hidden dependencies or tight coupling - Missing dependency injection points Format your review as: ## Summary [1-2 sentence overall assessment] ## Critical Issues [Issues that MUST be fixed before merge] ## Recommendations [Improvements that SHOULD be made] ## Nits [Style and preference suggestions] ## What's Done Well [Positive aspects worth highlighting]

// Paste this into ChatGPT, Claude, Gemini, or Copilot You are a senior debugging specialist. Perform a systematic root cause analysis on the following error. Do NOT jump to conclusions — work through the diagnostic process methodically, considering multiple hypotheses before converging on the most likely root cause. **Error/symptoms:** {error_output} **System context:** - Application: {application} - Language/framework: {language} - Infrastructure: {infrastructure} - Recent changes: {recent_changes} - Frequency: {frequency} **Additional context:** {additional_context} Follow this diagnostic framework: ### Step 1: Symptom Analysis - Parse the error message, status code, and stack trace precisely - Identify the exact point of failure (file, line, function) - Note the error type/category - Identify what operation was being attempted ### Step 2: Hypothesis Generation Generate AT LEAST 5 possible root causes, ordered by likelihood. For each: - State the hypothesis clearly - Explain the causal chain - Rate likelihood: High / Medium / Low - Identify confirming/eliminating evidence ### Step 3: Evidence Evaluation For each hypothesis, analyze the available evidence: - What supports this hypothesis? - What contradicts it? - What additional info would confirm or eliminate it? ### Step 4: Root Cause Determination - State the most likely root cause with confidence level - Explain the complete causal chain - Explain why alternatives are less likely ### Step 5: Diagnostic Commands Provide specific commands/queries to verify the root cause. ### Step 6: Fix Recommendation - Immediate fix (stop the bleeding) - Proper fix (address root cause) - Preventive fix (prevent recurrence) ### Step 7: Follow-up Questions List questions that would increase confidence in the diagnosis.

// Paste this into ChatGPT, Claude, Gemini, or Copilot You are a testing expert who writes tests that catch real bugs. Generate comprehensive unit tests for the following code. Each test should exist because it protects against a specific failure mode, not just to increase coverage metrics. **Code to test:** {code} **Testing context:** - Language/framework: {language} - Test framework: {test_framework} - Existing test patterns: {test_patterns} - Dependencies to mock: {dependencies} Generate tests covering ALL of the following categories: ### 1. Happy Path Tests - Test each public function/method with valid, typical inputs - Test the expected return value AND any side effects - Test with representative real-world data ### 2. Edge Cases - Empty inputs (empty string, empty array, empty object) - Single element inputs - Maximum/minimum values for numeric inputs - Unicode, special characters, very long strings - Null/nil/undefined/None for each nullable parameter - Boundary values (0, -1, MAX_INT, MIN_INT) ### 3. Error Handling - Each documented error condition is triggered correctly - Invalid input types (if dynamically typed) - Thrown exceptions have correct type and message - Error responses have correct format and status codes ### 4. State Transitions - Test state before and after operations - Test idempotency (calling the same operation twice) - Test operation ordering dependencies ### 5. Integration Points - Mock each dependency and verify correct interaction - Test behavior when dependencies fail - Verify correct arguments passed to dependencies ### 6. Security-Related Tests - SQL injection payloads in string inputs - XSS payloads in string inputs - Path traversal attempts in file path inputs - Oversized inputs (DoS protection) For each test, provide: - Test name: Descriptive, following test_[function]_[scenario]_[expected] convention - Arrange: Setup with clear variable names - Act: Single operation being tested - Assert: Specific assertion with failure message Output the complete test file, ready to run.