Menü
Consolidated master domain for prompt engineering.
Mit Codex oder Claude installieren Kopieren Sie diesen Prompt, fügen Sie ihn in Codex, Claude oder einen anderen Assistant ein und lassen Sie die Skill-Seite prüfen und installieren.
Basierend auf der SOC-Berufsklassifikation
Consolidated master domain for agentic arch.
Recognize patterns of context failure: lost-in-middle, poisoning, distraction, and clash
Build AI agents that interact with computers like humans do - viewing screens, moving cursors, clicking buttons, and typing text. Covers Anthropic's Computer Use, OpenAI's Operator/CUA, and open-source alternatives. Critical focus on sandboxing, security, and handling the unique challenges of vision-based control. Use when: computer use, desktop automation agent, screen control AI, vision-based agent, GUI automation.
Implement comprehensive evaluation strategies for LLM applications using automated metrics, human feedback, and benchmarking. Use when testing LLM performance, measuring AI application quality, or establishing evaluation frameworks.
Production-ready patterns for building LLM applications. Covers RAG pipelines, agent architectures, prompt IDEs, and LLMOps monitoring. Use when designing AI applications, implementing RAG, building agents, or setting up LLM observability.
Design short-term, long-term, and graph-based memory architectures
| name | prompt-engineering |
| description | Consolidated master domain for prompt engineering. |
| version | 1 |
| author | Antigravity |
[!IMPORTANT] This skill MUST be executed strictly under the Omni-Architect Agent Protocol v1.0. All tool executions, code modifications, and communications MUST adhere to the 13 core protocols.
Advanced prompt engineering techniques to maximize LLM performance, reliability, and controllability.
Teach the model by showing examples instead of explaining rules. Include 2-5 input-output pairs that demonstrate the desired behavior. Use when you need consistent formatting, specific reasoning patterns, or handling of edge cases. More examples improve accuracy but consume tokens—balance based on task complexity.
Example:
Extract key information from support tickets:
Input: "My login doesn't work and I keep getting error 403"
Output: {"issue": "authentication", "error_code": "403", "priority": "high"}
Input: "Feature request: add dark mode to settings"
Output: {"issue": "feature_request", "error_code": null, "priority": "low"}
Now process: "Can't upload files larger than 10MB, getting timeout"
Request step-by-step reasoning before the final answer. Add "Let's think step by step" (zero-shot) or include example reasoning traces (few-shot). Use for complex problems requiring multi-step logic, mathematical reasoning, or when you need to verify the model's thought process. Improves accuracy on analytical tasks by 30-50%.
Example:
Analyze this bug report and determine root cause.
Think step by step:
1. What is the expected behavior?
2. What is the actual behavior?
3. What changed recently that could cause this?
4. What components are involved?
5. What is the most likely root cause?
Bug: "Users can't save drafts after the cache update deployed yesterday"
Systematically improve prompts through testing and refinement. Start simple, measure performance (accuracy, consistency, token usage), then iterate. Test on diverse inputs including edge cases. Use A/B testing to compare variations. Critical for production prompts where consistency and cost matter.
Example:
Version 1 (Simple): "Summarize this article"
→ Result: Inconsistent length, misses key points
Version 2 (Add constraints): "Summarize in 3 bullet points"
→ Result: Better structure, but still misses nuance
Version 3 (Add reasoning): "Identify the 3 main findings, then summarize each"
→ Result: Consistent, accurate, captures key information
Build reusable prompt structures with variables, conditional sections, and modular components. Use for multi-turn conversations, role-based interactions, or when the same pattern applies to different inputs. Reduces duplication and ensures consistency across similar tasks.
Example:
# Reusable code review template
template = """
Review this {language} code for {focus_area}.
Code:
{code_block}
Provide feedback on:
{checklist}
"""
# Usage
prompt = template.format(
language="Python",
focus_area="security vulnerabilities",
code_block=user_code,
checklist="1. SQL injection\n2. XSS risks\n3. Authentication"
)
Set global behavior and constraints that persist across the conversation. Define the model's role, expertise level, output format, and safety guidelines. Use system prompts for stable instructions that shouldn't change turn-to-turn, freeing up user message tokens for variable content.
Example:
System: You are a senior backend engineer specializing in API design.
Rules:
- Always consider scalability and performance
- Suggest RESTful patterns by default
- Flag security concerns immediately
- Provide code examples in Python
- Use early return pattern
Format responses as:
1. Analysis
2. Recommendation
3. Code example
4. Trade-offs
Start with simple prompts, add complexity only when needed:
Level 1: Direct instruction
Level 2: Add constraints
Level 3: Add reasoning
Level 4: Add examples
[System Context] → [Task Instruction] → [Examples] → [Input Data] → [Output Format]
Build prompts that gracefully handle failures:
A comprehensive collection of battle-tested prompts inspired by awesome-chatgpt-prompts and community best practices.
Use this skill when the user:
Act as an expert software developer with 15+ years of experience. You specialize in clean code, SOLID principles, and pragmatic architecture. When reviewing code:
1. Identify bugs and potential issues
2. Suggest performance improvements
3. Recommend better patterns
4. Explain your reasoning clearly
Always prioritize readability and maintainability over cleverness.
Act as a senior code reviewer. Your role is to:
1. Check for bugs, edge cases, and error handling
2. Evaluate code structure and organization
3. Assess naming conventions and readability
4. Identify potential security issues
5. Suggest improvements with specific examples
Format your review as:
🔴 Critical Issues (must fix)
🟡 Suggestions (should consider)
🟢 Praise (what's done well)
Act as a technical documentation expert. Transform complex technical concepts into clear, accessible documentation. Follow these principles:
- Use simple language, avoid jargon
- Include practical examples
- Structure with clear headings
- Add code snippets where helpful
- Consider the reader's experience level
Act as a senior system architect designing for scale. Consider:
- Scalability (horizontal and vertical)
- Reliability (fault tolerance, redundancy)
- Maintainability (modularity, clear boundaries)
- Performance (latency, throughput)
- Cost efficiency
Provide architecture decisions with trade-off analysis.
Debug the following code. Your analysis should include:
1. **Problem Identification**: What exactly is failing?
2. **Root Cause**: Why is it failing?
3. **Fix**: Provide corrected code
4. **Prevention**: How to prevent similar bugs
Show your debugging thought process step by step.
Explain [CONCEPT] as if I'm 5 years old. Use:
- Simple everyday analogies
- No technical jargon
- Short sentences
- Relatable examples from daily life
- A fun, engaging tone
Refactor this code following these priorities:
1. Readability first
2. Remove duplication (DRY)
3. Single responsibility per function
4. Meaningful names
5. Add comments only where necessary
Show before/after with explanation of changes.
Write comprehensive tests for this code:
1. Happy path scenarios
2. Edge cases
3. Error conditions
4. Boundary values
Use [FRAMEWORK] testing conventions. Include:
- Descriptive test names
- Arrange-Act-Assert pattern
- Mocking where appropriate
Generate API documentation for this endpoint including:
- Endpoint URL and method
- Request parameters (path, query, body)
- Request/response examples
- Error codes and meanings
- Authentication requirements
- Rate limits if applicable
Format as OpenAPI/Swagger or Markdown.
Analyze the complexity of this codebase:
1. **Cyclomatic Complexity**: Identify complex functions
2. **Coupling**: Find tightly coupled components
3. **Cohesion**: Assess module cohesion
4. **Dependencies**: Map critical dependencies
5. **Technical Debt**: Highlight areas needing refactoring
Rate each area and provide actionable recommendations.
Analyze this code for performance issues:
1. **Time Complexity**: Big O analysis
2. **Space Complexity**: Memory usage patterns
3. **I/O Bottlenecks**: Database, network, disk
4. **Algorithmic Issues**: Inefficient patterns
5. **Quick Wins**: Easy optimizations
Prioritize findings by impact.
Perform a security review of this code:
1. **Input Validation**: Check all inputs
2. **Authentication/Authorization**: Access control
3. **Data Protection**: Sensitive data handling
4. **Injection Vulnerabilities**: SQL, XSS, etc.
5. **Dependencies**: Known vulnerabilities
Classify issues by severity (Critical/High/Medium/Low).
Brainstorm features for [PRODUCT]:
For each feature, provide:
- Name and one-line description
- User value proposition
- Implementation complexity (Low/Med/High)
- Dependencies on other features
Generate 10 ideas, then rank top 3 by impact/effort ratio.
Generate names for [PROJECT/FEATURE]:
Provide 10 options in these categories:
- Descriptive (what it does)
- Evocative (how it feels)
- Acronyms (memorable abbreviations)
- Metaphorical (analogies)
For each, explain the reasoning and check domain availability patterns.
Migrate this code from [SOURCE] to [TARGET]:
1. Identify equivalent constructs
2. Handle incompatible features
3. Preserve functionality exactly
4. Follow target language idioms
5. Add necessary dependencies
Show the migration step by step with explanations.
Convert this [SOURCE_FORMAT] to [TARGET_FORMAT]:
Requirements:
- Preserve all data
- Use idiomatic target format
- Handle edge cases
- Validate the output
- Provide sample verification
Let's solve this step by step:
1. First, I'll understand the problem
2. Then, I'll identify the key components
3. Next, I'll work through the logic
4. Finally, I'll verify the solution
[Your question here]
Here are some examples of the task:
Example 1:
Input: [example input 1]
Output: [example output 1]
Example 2:
Input: [example input 2]
Output: [example output 2]
Now complete this:
Input: [actual input]
Output:
You are [PERSONA] with [TRAITS].
Your communication style is [STYLE].
You prioritize [VALUES].
When responding:
- [Behavior 1]
- [Behavior 2]
- [Behavior 3]
Respond in the following JSON format:
{
"analysis": "your analysis here",
"recommendations": ["rec1", "rec2"],
"confidence": 0.0-1.0,
"caveats": ["caveat1"]
}
When crafting prompts, ensure:
💡 Tip: The best prompts are specific, provide context, and include examples of desired output.
Version: 1.0.1
Status: ✨ Zero-Config | 🌍 Universal
Transform raw prompts into optimized, production-ready prompts using 11 established prompting frameworks.
Prompt Engineer is an intelligent AI skill that analyzes your intentions and automatically generates optimized prompts for Claude, ChatGPT, or any other AI model. Instead of struggling with how to phrase complex requests, simply describe what you want - the skill handles the rest.
This skill works in "magic mode" - it operates silently, only asking questions when absolutely necessary. You provide a rough idea, and it returns a polished, structured prompt ready to use.
Use any of these trigger phrases:
copilot> improve this prompt: [your rough prompt]
copilot> optimize prompt: [your idea]
copilot> refine prompt: [what you want]
copilot> prompt engineering for: [task description]
copilot> transform into prompt: [concept]
copilot> create prompt for: [goal]
Input:
copilot> improve this prompt: help me write Python code
Skill asks:
What type of Python code do you need?
- [ ] Data analysis
- [ ] Web scraping
- [ ] Task automation
- [ ] API/backend development
You select: "Web scraping"
Output:
```
You are a Python developer expert in web scraping.
Task: Create a Python script to scrape data from [specify target website].
Requirements:
- Use BeautifulSoup or Scrapy
- Implement pagination and rate limiting
- Store data in structured format (CSV or JSON)
- Include error handling for network issues
Output format: Complete, commented Python code with usage examples.
```
Input:
copilot> create prompt for: product recommendation system with purchase history and market trends
Output:
```
Role: You are a senior machine learning engineer specializing in recommendation systems.
Objective: Design a product recommendation system that combines user purchase history, preferences, and market trends.
Details:
- Input data sources: user transaction logs, preference surveys, market trend APIs
- Algorithm considerations: collaborative filtering, content-based filtering, hybrid approaches
- Performance metrics: precision, recall, diversity, novelty
- Scalability: system must handle 100k+ users
Approach step-by-step:
1. Analyze the three data sources and their relationships
2. Propose 2-3 algorithm approaches with trade-offs
3. Design data pipeline architecture
4. Define evaluation metrics
5. Outline implementation roadmap
Examples: Show pseudocode for core recommendation logic
Sense Check: Validate that the solution balances personalization with discovery and scales to production.
Output format: Technical design document with architecture diagrams and implementation roadmap.
```
Input:
copilot> transform into prompt: I want to improve my process
Skill asks:
What process do you want to improve?
- [ ] Work/productivity process
- [ ] Learning process
- [ ] Technical process (development, deployment, etc.)
You select: "Work process"
Skill asks:
What specific aspect?
- [ ] Time management
- [ ] Team collaboration
- [ ] Decision making
You select: "Decision making"
Output:
```
Context: I struggle with making decisions quickly and effectively at work.
Objective: Develop a structured framework to improve decision speed and quality.
Current Reality:
- Decisions delayed by excessive analysis
- Lack of clear criteria
- Uncertainty about when to decide vs. gather information
Desired Outcome:
- Consistent decision-making framework
- Clear criteria for different decision types
- Reduced decision fatigue
Your task:
1. Assess current patterns and identify bottlenecks
2. Propose framework (Eisenhower Matrix, RAPID, Decision Tree)
3. Create decision criteria templates
4. Design tracking and iteration process
Output format: Practical playbook with templates, examples, and metrics.
```
The Prompt Engineer skill uses 11 established frameworks to optimize your prompts. The skill automatically selects and blends these frameworks based on your task - you never need to know or choose them manually.
Structure: Role → Task → Format
Best for: Tasks requiring specific expertise or perspective
Components:
Example:
You are a senior Python developer.
Task: Refactor this code for better performance.
Format: Provide refactored code with inline comments explaining changes.
Structure: Problem → Step 1 → Step 2 → ... → Solution
Best for: Complex reasoning, debugging, mathematical problems, logic puzzles
Components:
Example:
Solve this problem step-by-step:
1. Identify the core issue
2. Analyze contributing factors
3. Propose solution approach
4. Validate solution against requirements
Structure: Role, Instructions, Steps, End goal, Narrowing
Best for: Multi-phase projects with clear deliverables and constraints
Components:
Example:
Role: You are a DevOps architect.
Instructions: Design a CI/CD pipeline for microservices.
Steps: 1) Analyze requirements 2) Select tools 3) Design workflow 4) Document
End goal: Automated deployment with zero-downtime releases.
Narrowing: Focus on AWS, limit to 3 environments (dev/staging/prod).
Structure: Role, Objective, Details, Examples, Sense check
Best for: Complex design, system architecture, research proposals
Components:
Example:
Role: You are a system architect.
Objective: Design a scalable e-commerce platform.
Details: Handle 100k concurrent users, sub-200ms response time, multi-region.
Examples: Show database schema, caching strategy, load balancing.
Sense check: Validate solution meets latency and scalability requirements.
Structure: Iteration 1 (verbose) → Iteration 2 → ... → Iteration 5 (maximum density)
Best for: Summarization, compression, synthesis of long content
Process:
Example:
Compress this article into progressively denser summaries:
1. Initial summary (300 words)
2. Compressed (200 words)
3. Further compressed (100 words)
4. Dense (50 words)
5. Maximum density (25 words, all critical points)
Structure: Role, Audience, Context, Expectation
Best for: Communication, presentations, stakeholder updates, storytelling
Components:
Example:
Role: You are a product manager.
Audience: Non-technical executives.
Context: Quarterly business review, product performance down 5%.
Expectation: Explain root causes and recovery plan in non-technical terms.
Structure: Research, Investigate, Synthesize, Evaluate
Best for: Analysis, investigation, systematic exploration, diagnostic work
Process:
Example:
Analyze customer churn data using RISE:
Research: Collect churn metrics, exit surveys, support tickets.
Investigate: Identify patterns in churned users.
Synthesize: Combine findings into themes.
Evaluate: Recommend retention strategies based on evidence.
Structure: Situation, Task, Action, Result
Best for: Problem-solving with rich context, case studies, retrospectives
Components:
Example:
Situation: Legacy monolith causing deployment delays (2 weeks per release).
Task: Modernize architecture to enable daily deployments.
Action: Migrate to microservices, implement CI/CD, containerize.
Result: Deploy 10+ times per day with <5% rollback rate.
Structure: Subjective, Objective, Assessment, Plan
Best for: Structured documentation, medical records, technical logs, incident reports
Components:
Example:
Incident Report (SOAP):
Subjective: Users report slow page loads starting 10 AM.
Objective: Average response time increased from 200ms to 3s. CPU at 95%.
Assessment: Database connection pool exhausted due to traffic spike.
Plan: 1) Scale pool size 2) Add monitoring alerts 3) Review query performance.
Structure: Collaborative, Limited, Emotional, Appreciable, Refinable
Best for: Goal-setting, OKRs, measurable objectives, team alignment
Components:
Example:
Q1 Objective (CLEAR):
Collaborative: Engineering + Product teams.
Limited: Complete by March 31, budget $50k, 2 engineers allocated.
Emotional: Reduces customer support load by 30%, improves satisfaction.
Appreciable: Track weekly via tickets resolved, NPS score, deployment count.
Refinable: Bi-weekly retrospectives, adjust priorities based on feedback.
Structure: Goal, Reality, Options, Will
Best for: Coaching, personal development, growth planning, mentorship
Components:
Example:
Career Development (GROW):
Goal: Become senior engineer within 12 months.
Reality: Strong coding skills, weak in system design and leadership.
Options: 1) Take system design course 2) Lead a project 3) Find mentor.
Will: Commit to 5 hours/week study, lead Q2 project, find mentor by Feb.
The skill analyzes your input and:
Detects task type
Identifies complexity
Selects primary framework
Blends secondary frameworks when needed
You never choose the framework manually - the skill does it automatically in "magic mode."
| Task Type | Primary Framework | Blended With | Result |
|---|---|---|---|
| Complex technical design | RODES | Chain of Thought | Structured design with step-by-step reasoning |
| Leadership development | CLEAR | GROW | Measurable goals with action commitment |
| Strategic communication | RACE | STAR | Audience-aware storytelling with context |
| Incident investigation | RISE | SOAP | Systematic analysis with structured documentation |
| Project planning | RISEN | RTF | Multi-phase delivery with role clarity |
User Input (rough prompt)
↓
┌────────────────────────┐
│ 1. Analyze Intent │ What is the user trying to do?
│ - Task type │ Coding? Writing? Analysis? Design?
│ - Complexity │ Simple, moderate, complex?
│ - Clarity │ Clear or ambiguous?
└────────┬───────────────┘
↓
┌────────────────────────┐
│ 2. Clarify (Optional) │ Only if critically needed
│ - Ask 2-3 questions │ Multiple choice when possible
│ - Fill missing gaps │
└────────┬───────────────┘
↓
┌────────────────────────┐
│ 3. Select Framework(s) │ Silent selection
│ - Map task → framework
│ - Blend if needed │
└────────┬───────────────┘
↓
┌────────────────────────┐
│ 4. Generate Prompt │ Apply framework rules
│ - Add role/context │
│ - Structure task │
│ - Define format │
│ - Add examples │
└────────┬───────────────┘
↓
┌────────────────────────┐
│ 5. Output │ Clean, copy-ready
│ Markdown code block │ No explanations
└────────────────────────┘
copilot> optimize prompt: create REST API in Python
→ Generates structured prompt with role, requirements, output format, examples
copilot> create prompt for: write technical article about microservices
→ Generates audience-aware prompt with structure, tone, and content guidelines
copilot> refine prompt: analyze sales data and identify trends
→ Generates step-by-step analytical framework with visualization requirements
copilot> improve this prompt: I need to decide between technology A and B
→ Generates decision framework with criteria, trade-offs, and validation
copilot> transform into prompt: learn machine learning from zero
→ Generates learning path prompt with phases, resources, and milestones
A: Yes! It's a universal skill that works in any terminal context. It doesn't depend on vault structure, project configuration, or external files.
A: No. The skill knows all 11 frameworks and selects the best one(s) automatically based on your task.
A: No. It operates in "magic mode" - you get the polished prompt without technical explanations. If you want to know, you can ask explicitly.
A: Maximum 2-3 questions, and only when information is critically missing. Most of the time, it generates the prompt directly.
A: The skill uses standard framework definitions. You can't customize them, but you can provide additional constraints in your input (e.g., "create a short prompt for...").
A: Yes. If you provide input in Portuguese, it generates the prompt in Portuguese. Same for English or mixed inputs.
A: You can ask the skill to refine it: "make it shorter", "add more examples", "focus on X aspect", etc.
A: Yes. The prompts are model-agnostic and work with any conversational AI.
This skill is designed to work globally across all your projects.
Clone the repository:
git clone https://github.com/eric.andrade/cli-ai-skills.git
Configure Copilot to load skills globally:
# Add to ~/.copilot/config.json
{
"skills": {
"directories": [
"/path/to/cli-ai-skills/.github/skills"
]
}
}
cp -r /path/to/cli-ai-skills/.github/skills/prompt-engineer ~/.copilot/global-skills/
Then configure:
# Add to ~/.copilot/config.json
{
"skills": {
"directories": [
"~/.copilot/global-skills"
]
}
}
v1.0.1 | Zero-Config | Universal
Works in any project, any context, any terminal.
This skill transforms raw, unstructured user prompts into highly optimized prompts using established prompting frameworks. It analyzes user intent, identifies task complexity, and intelligently selects the most appropriate framework(s) to maximize Claude/ChatGPT output quality.
The skill operates in "magic mode" - it works silently behind the scenes, only interacting with users when clarification is critically needed. Users receive polished, ready-to-use prompts without technical explanations or framework jargon.
This is a universal skill that works in any terminal context, not limited to Obsidian vaults or specific project structures.
Invoke this skill when:
Objective: Understand what the user truly wants to accomplish.
Actions:
Detection Patterns:
Objective: Map task characteristics to optimal prompting framework(s).
Framework Mapping Logic:
| Task Type | Recommended Framework(s) | Rationale |
|---|---|---|
| Role-based tasks (act as expert, consultant) | RTF (Role-Task-Format) | Clear role definition + task + output format |
| Step-by-step reasoning (debugging, proof, logic) | Chain of Thought | Encourages explicit reasoning steps |
| Structured projects (multi-phase, deliverables) | RISEN (Role, Instructions, Steps, End goal, Narrowing) | Comprehensive structure for complex work |
| Complex design/analysis (systems, architecture) | RODES (Role, Objective, Details, Examples, Sense check) | Balances detail with validation |
| Summarization (compress, synthesize) | Chain of Density | Iterative refinement to essential info |
| Communication (reports, presentations, storytelling) | RACE (Role, Audience, Context, Expectation) | Audience-aware messaging |
| Investigation/analysis (research, diagnosis) | RISE (Research, Investigate, Synthesize, Evaluate) | Systematic analytical approach |
| Contextual situations (problem-solving with background) | STAR (Situation, Task, Action, Result) | Context-rich problem framing |
| Documentation (medical, technical, records) | SOAP (Subjective, Objective, Assessment, Plan) | Structured information capture |
| Goal-setting (OKRs, objectives, targets) | CLEAR (Collaborative, Limited, Emotional, Appreciable, Refinable) | Goal clarity and actionability |
| Coaching/development (mentoring, growth) | GROW (Goal, Reality, Options, Will) | Developmental conversation structure |
Blending Strategy:
Selection Criteria:
Critical Rule: This selection happens silently - do not explain framework choice to user.
Role: You are a senior software architect. [RTF - Role]
Objective: Design a microservices architecture for [system]. [RODES - Objective]
Approach this step-by-step: [Chain of Thought]
Details: [RODES - Details]
Output Format: [RTF - Format] Provide architecture diagram description, service definitions, and migration roadmap.
Sense Check: [RODES - Sense check] Validate that services are loosely coupled, independently deployable, and aligned with business domains.
**4.5. Language Adaptation**
- If original prompt is in Portuguese, generate prompt in Portuguese
- If original prompt is in English, generate prompt in English
- If mixed, default to English (more universal for AI models)
**4.6. Quality Checks**
Before finalizing, verify:
- [ ] Prompt is self-contained (no external context needed)
- [ ] Task is specific and measurable
- [ ] Output format is clear
- [ ] No ambiguous language
- [ ] Appropriate level of detail for task complexity
## Critical Rules
### **NEVER:**
- ❌ Assume information that wasn't provided - ALWAYS ask if critical details are missing
- ❌ Explain which framework was selected or why (magic mode - keep it invisible)
- ❌ Generate generic, one-size-fits-all prompts - always customize to context
- ❌ Use technical jargon in the final prompt (unless user's domain is technical)
- ❌ Ask more than 3 clarifying questions (avoid user fatigue)
- ❌ Include meta-commentary in the output ("This prompt uses...", "Note that...")
- ❌ Present output without code block formatting
- ❌ Mix languages inconsistently (if user writes in PT, respond in PT)
### **ALWAYS:**
- ✅ Analyze intent before generating (Step 1 is mandatory)
- ✅ Ask clarifying questions if critical information is ambiguous (Step 2 conditional)
- ✅ Select framework(s) based on task type and complexity (Step 3 mapping)
- ✅ Blend multiple frameworks when it improves prompt quality
- ✅ Adapt prompt length to original input complexity (simple → short, complex → detailed)
- ✅ Include output format specification in generated prompts
- ✅ Present final prompt in clean Markdown code block
- ✅ Make prompts self-contained (no dependency on external context)
- ✅ Use examples in complex prompts to illustrate expected output
- ✅ Validate prompt completeness before presenting (quality checks in Step 4.6)
### **Example 2: Complex Clear Prompt → RODES + Chain of Thought (No Clarification)**
**User Input:**
copilot> create prompt for: I need to build a product recommendation system considering purchase history, user preferences, and market trends
**Skill Internal Process:**
1. **Analyze:** Clear intent, complex task (ML system design), technical domain
2. **Clarify:** Skip (sufficient information)
3. **Select:** RODES (complex design) + Chain of Thought (reasoning needed)
**Skill Output:**
````markdown
Role: You are a senior machine learning engineer specializing in recommendation systems.
Objective: Design a product recommendation system that combines user purchase history, preferences, and market trends.
Details:
Approach step-by-step:
Examples:
Sense Check: Validate that the proposed solution balances personalization with discovery, scales to production load, and incorporates real-time market trends effectively.
Output format: Technical design document with architecture diagrams, algorithm justification, and implementation roadmap.
User Input:
copilot> optimize prompt: my Python code isn't working, needs debugging
Skill Internal Process:
Skill Output:
```
You are an expert Python debugger.
Task: Debug the Python code provided by the user by analyzing it step-by-step.
Debugging approach:
1. Read and understand the code's intended purpose
2. Identify the error or unexpected behavior
3. Trace variable states at each execution step
4. Locate the root cause of the issue
5. Propose fix with explanation
6. Suggest preventive measures to avoid similar bugs
For each step, show your reasoning:
- What you're checking
- What you found
- Why it matters
Output format:
- **Issue identified:** [Describe the bug]
- **Root cause:** [Why it's happening]
- **Fix:** [Corrected code with comments]
- **Prevention:** [Best practices to avoid this in the future]
Include a working example to verify the fix.
```
This skill is platform-agnostic and works in any terminal context where GitHub Copilot CLI is available. It does not depend on:
The skill is entirely self-contained, operating purely on user input and framework knowledge.
Advanced prompt engineering techniques to maximize LLM performance, reliability, and controllability.
Teach the model by showing examples instead of explaining rules. Include 2-5 input-output pairs that demonstrate the desired behavior. Use when you need consistent formatting, specific reasoning patterns, or handling of edge cases. More examples improve accuracy but consume tokens—balance based on task complexity.
Example:
Extract key information from support tickets:
Input: "My login doesn't work and I keep getting error 403"
Output: {"issue": "authentication", "error_code": "403", "priority": "high"}
Input: "Feature request: add dark mode to settings"
Output: {"issue": "feature_request", "error_code": null, "priority": "low"}
Now process: "Can't upload files larger than 10MB, getting timeout"
Request step-by-step reasoning before the final answer. Add "Let's think step by step" (zero-shot) or include example reasoning traces (few-shot). Use for complex problems requiring multi-step logic, mathematical reasoning, or when you need to verify the model's thought process. Improves accuracy on analytical tasks by 30-50%.
Example:
Analyze this bug report and determine root cause.
Think step by step:
1. What is the expected behavior?
2. What is the actual behavior?
3. What changed recently that could cause this?
4. What components are involved?
5. What is the most likely root cause?
Bug: "Users can't save drafts after the cache update deployed yesterday"
Systematically improve prompts through testing and refinement. Start simple, measure performance (accuracy, consistency, token usage), then iterate. Test on diverse inputs including edge cases. Use A/B testing to compare variations. Critical for production prompts where consistency and cost matter.
Example:
Version 1 (Simple): "Summarize this article"
→ Result: Inconsistent length, misses key points
Version 2 (Add constraints): "Summarize in 3 bullet points"
→ Result: Better structure, but still misses nuance
Version 3 (Add reasoning): "Identify the 3 main findings, then summarize each"
→ Result: Consistent, accurate, captures key information
Build reusable prompt structures with variables, conditional sections, and modular components. Use for multi-turn conversations, role-based interactions, or when the same pattern applies to different inputs. Reduces duplication and ensures consistency across similar tasks.
Example:
# Reusable code review template
template = """
Review this {language} code for {focus_area}.
Code:
{code_block}
Provide feedback on:
{checklist}
"""
# Usage
prompt = template.format(
language="Python",
focus_area="security vulnerabilities",
code_block=user_code,
checklist="1. SQL injection\n2. XSS risks\n3. Authentication"
)
Set global behavior and constraints that persist across the conversation. Define the model's role, expertise level, output format, and safety guidelines. Use system prompts for stable instructions that shouldn't change turn-to-turn, freeing up user message tokens for variable content.
Example:
System: You are a senior backend engineer specializing in API design.
Rules:
- Always consider scalability and performance
- Suggest RESTful patterns by default
- Flag security concerns immediately
- Provide code examples in Python
- Use early return pattern
Format responses as:
1. Analysis
2. Recommendation
3. Code example
4. Trade-offs
Start with simple prompts, add complexity only when needed:
Level 1: Direct instruction
Level 2: Add constraints
Level 3: Add reasoning
Level 4: Add examples
[System Context] → [Task Instruction] → [Examples] → [Input Data] → [Output Format]
Build prompts that gracefully handle failures:
A comprehensive collection of battle-tested prompts inspired by awesome-chatgpt-prompts and community best practices.
Use this skill when the user:
Act as an expert software developer with 15+ years of experience. You specialize in clean code, SOLID principles, and pragmatic architecture. When reviewing code:
1. Identify bugs and potential issues
2. Suggest performance improvements
3. Recommend better patterns
4. Explain your reasoning clearly
Always prioritize readability and maintainability over cleverness.
Act as a senior code reviewer. Your role is to:
1. Check for bugs, edge cases, and error handling
2. Evaluate code structure and organization
3. Assess naming conventions and readability
4. Identify potential security issues
5. Suggest improvements with specific examples
Format your review as:
🔴 Critical Issues (must fix)
🟡 Suggestions (should consider)
🟢 Praise (what's done well)
Act as a technical documentation expert. Transform complex technical concepts into clear, accessible documentation. Follow these principles:
- Use simple language, avoid jargon
- Include practical examples
- Structure with clear headings
- Add code snippets where helpful
- Consider the reader's experience level
Act as a senior system architect designing for scale. Consider:
- Scalability (horizontal and vertical)
- Reliability (fault tolerance, redundancy)
- Maintainability (modularity, clear boundaries)
- Performance (latency, throughput)
- Cost efficiency
Provide architecture decisions with trade-off analysis.
Debug the following code. Your analysis should include:
1. **Problem Identification**: What exactly is failing?
2. **Root Cause**: Why is it failing?
3. **Fix**: Provide corrected code
4. **Prevention**: How to prevent similar bugs
Show your debugging thought process step by step.
Explain [CONCEPT] as if I'm 5 years old. Use:
- Simple everyday analogies
- No technical jargon
- Short sentences
- Relatable examples from daily life
- A fun, engaging tone
Refactor this code following these priorities:
1. Readability first
2. Remove duplication (DRY)
3. Single responsibility per function
4. Meaningful names
5. Add comments only where necessary
Show before/after with explanation of changes.
Write comprehensive tests for this code:
1. Happy path scenarios
2. Edge cases
3. Error conditions
4. Boundary values
Use [FRAMEWORK] testing conventions. Include:
- Descriptive test names
- Arrange-Act-Assert pattern
- Mocking where appropriate
Generate API documentation for this endpoint including:
- Endpoint URL and method
- Request parameters (path, query, body)
- Request/response examples
- Error codes and meanings
- Authentication requirements
- Rate limits if applicable
Format as OpenAPI/Swagger or Markdown.
Analyze the complexity of this codebase:
1. **Cyclomatic Complexity**: Identify complex functions
2. **Coupling**: Find tightly coupled components
3. **Cohesion**: Assess module cohesion
4. **Dependencies**: Map critical dependencies
5. **Technical Debt**: Highlight areas needing refactoring
Rate each area and provide actionable recommendations.
Analyze this code for performance issues:
1. **Time Complexity**: Big O analysis
2. **Space Complexity**: Memory usage patterns
3. **I/O Bottlenecks**: Database, network, disk
4. **Algorithmic Issues**: Inefficient patterns
5. **Quick Wins**: Easy optimizations
Prioritize findings by impact.
Perform a security review of this code:
1. **Input Validation**: Check all inputs
2. **Authentication/Authorization**: Access control
3. **Data Protection**: Sensitive data handling
4. **Injection Vulnerabilities**: SQL, XSS, etc.
5. **Dependencies**: Known vulnerabilities
Classify issues by severity (Critical/High/Medium/Low).
Brainstorm features for [PRODUCT]:
For each feature, provide:
- Name and one-line description
- User value proposition
- Implementation complexity (Low/Med/High)
- Dependencies on other features
Generate 10 ideas, then rank top 3 by impact/effort ratio.
Generate names for [PROJECT/FEATURE]:
Provide 10 options in these categories:
- Descriptive (what it does)
- Evocative (how it feels)
- Acronyms (memorable abbreviations)
- Metaphorical (analogies)
For each, explain the reasoning and check domain availability patterns.
Migrate this code from [SOURCE] to [TARGET]:
1. Identify equivalent constructs
2. Handle incompatible features
3. Preserve functionality exactly
4. Follow target language idioms
5. Add necessary dependencies
Show the migration step by step with explanations.
Convert this [SOURCE_FORMAT] to [TARGET_FORMAT]:
Requirements:
- Preserve all data
- Use idiomatic target format
- Handle edge cases
- Validate the output
- Provide sample verification
Let's solve this step by step:
1. First, I'll understand the problem
2. Then, I'll identify the key components
3. Next, I'll work through the logic
4. Finally, I'll verify the solution
[Your question here]
Here are some examples of the task:
Example 1:
Input: [example input 1]
Output: [example output 1]
Example 2:
Input: [example input 2]
Output: [example output 2]
Now complete this:
Input: [actual input]
Output:
You are [PERSONA] with [TRAITS].
Your communication style is [STYLE].
You prioritize [VALUES].
When responding:
- [Behavior 1]
- [Behavior 2]
- [Behavior 3]
Respond in the following JSON format:
{
"analysis": "your analysis here",
"recommendations": ["rec1", "rec2"],
"confidence": 0.0-1.0,
"caveats": ["caveat1"]
}
When crafting prompts, ensure:
💡 Tip: The best prompts are specific, provide context, and include examples of desired output.
gemini-api-dev skill into antigravity-config.gemini-3-flash-preview, gemini-3-pro-preview).google-genai), JS/TS (@google/genai), Go, and Java.The Gemini API provides access to Google's most advanced AI models. Key capabilities include:
gemini-3-pro-preview: 1M tokens, complex reasoning, coding, researchgemini-3-flash-preview: 1M tokens, fast, balanced performance, multimodalgemini-3-pro-image-preview: 65k / 32k tokens, image generation and editing[!IMPORTANT] Models like
gemini-2.5-*,gemini-2.0-*,gemini-1.5-*are legacy and deprecated. Use the new models above. Your knowledge is outdated.
google-genai install with pip install google-genai@google/genai install with npm install @google/genaigoogle.golang.org/genai install with go get google.golang.org/genaicom.google.genai, artifactId: google-genaiLAST_VERSION)build.gradle:
implementation("com.google.genai:google-genai:${LAST_VERSION}")
pom.xml:
<dependency>
<groupId>com.google.genai</groupId>
<artifactId>google-genai</artifactId>
<version>${LAST_VERSION}</version>
</dependency>
[!WARNING] Legacy SDKs
google-generativeai(Python) and@google/generative-ai(JS) are deprecated. Migrate to the new SDKs above urgently by following the Migration Guide.
from google import genai
client = genai.Client()
response = client.models.generate_content(
model="gemini-3-flash-preview",
contents="Explain quantum computing"
)
print(response.text)
import { GoogleGenAI } from "@google/genai";
const ai = new GoogleGenAI({});
const response = await ai.models.generateContent({
model: "gemini-3-flash-preview",
contents: "Explain quantum computing"
});
console.log(response.text);
package main
import (
"context"
"fmt"
"log"
"google.golang.org/genai"
)
func main() {
ctx := context.Background()
client, err := genai.NewClient(ctx, nil)
if err != nil {
log.Fatal(err)
}
resp, err := client.Models.GenerateContent(ctx, "gemini-3-flash-preview", genai.Text("Explain quantum computing"), nil)
if err != nil {
log.Fatal(err)
}
fmt.Println(resp.Text)
}
import com.google.genai.Client;
import com.google.genai.types.GenerateContentResponse;
public class GenerateTextFromTextInput {
public static void main(String[] args) {
Client client = new Client();
GenerateContentResponse response =
client.models.generateContent(
"gemini-3-flash-preview",
"Explain quantum computing",
null);
System.out.println(response.text());
}
}
Always use the latest REST API discovery spec as the source of truth for API definitions (request/response schemas, parameters, methods). Fetch the spec when implementing or debugging API integration:
https://generativelanguage.googleapis.com/$discovery/rest?version=v1betahttps://generativelanguage.googleapis.com/$discovery/rest?version=v1When in doubt, use v1beta. Refer to the spec for exact field names, types, and supported operations.
For detailed API documentation, fetch from the official docs index:
llms.txt URL: https://ai.google.dev/gemini-api/docs/llms.txt
This index contains links to all documentation pages in .md.txt format. Use web fetch tools to:
llms.txt to discover available documentation pageshttps://ai.google.dev/gemini-api/docs/function-calling.md.txt)[!IMPORTANT] Those are not all the documentation pages. Use the
llms.txtindex to discover available documentation pages