Prompt engineering
Skill ComeOnOliver/skillshub/skills/aiskillstore/marketplace/doyajin174/prompt-engineering
π§ The right skill, one API call. AI agent skills registry with token-efficient skill resolution. 5,000+ skills from 500+ top repos.
npx -y skills add ComeOnOliver/skillshub --skill prompt-engineeringAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
What its author says it does
Copied from the file, not written here
Advanced prompt engineering techniques for optimal AI responses. Use this when crafting prompts, optimizing AI interactions, or designing system prompts for applications.
The file declares its own license as MIT. That is the authorβs claim about this one file, and it is not the same thing as the license GitHub reports for the repository, which is listed with the other numbers below.
SKILL.md
4.1 KB, as published. Nobody here has run it
Prompt Engineering
ν¨κ³Όμ μΈ ν둬ννΈ μμ±μ μν κ³ κΈ κΈ°λ² κ°μ΄λμ λλ€.
Core Principles
1. Clarity (λͺ νμ±)
β "μ½λ μ’ κ³ μ³μ€"
β
"auth.jsμ validateToken ν¨μμμ λ§λ£λ ν ν° μ²λ¦¬ μ
λ°μνλ TypeErrorλ₯Ό μμ ν΄μ£ΌμΈμ"
2. Context (λ§₯λ½)
β "μ΄ μλ¬ ν΄κ²°ν΄μ€: TypeError"
β
"Node.js 18, Express 4 νκ²½μμ JWT κ²μ¦ μ
λ€μ μλ¬ λ°μ: TypeError: Cannot read property 'exp' of undefined
κ΄λ ¨ μ½λ: [μ½λ 첨λΆ]"
3. Specificity (ꡬ체μ±)
β "μ’μ API λ§λ€μ΄μ€"
β
"λ€μ μꡬμ¬νμ REST API μ€κ³:
- 리μμ€: μ¬μ©μ, κ²μλ¬Ό
- μΈμ¦: JWT Bearer
- νμ΄μ§λ€μ΄μ
: cursor κΈ°λ°
- μλ΅ νμ: JSON:API μ€ν"
Prompting Techniques
Zero-Shot Prompting
μ§μ μ§μλ§μΌλ‘ κ²°κ³Ό λμΆ
μ: "λ€μ ν
μ€νΈμ κ°μ μ λΆμνμΈμ: [ν
μ€νΈ]"
Few-Shot Prompting
μμλ₯Ό ν΅ν ν¨ν΄ νμ΅
μμ 1: μ
λ ₯ β μΆλ ₯
μμ 2: μ
λ ₯ β μΆλ ₯
---
μ€μ μ
λ ₯ β ?
Chain-of-Thought (CoT)
λ¨κ³λ³ μΆλ‘ μ λ
"λ¨κ³λ³λ‘ μκ°ν΄λ³΄μΈμ:
1. λ¨Όμ λ¬Έμ λ₯Ό λΆμν©λλ€
2. κ°λ₯ν ν΄κ²°μ±
μ λμ΄ν©λλ€
3. κ° ν΄κ²°μ±
μ μ₯λ¨μ μ νκ°ν©λλ€
4. μ΅μ μ ν΄κ²°μ±
μ μ νν©λλ€"
Role Prompting
νΉμ μν λΆμ¬
"λΉμ μ 10λ
κ²½λ ₯μ μλμ΄ λ°±μλ κ°λ°μμ
λλ€.
μ½λ 리뷰 κ΄μ μμ λ€μ μ½λλ₯Ό νκ°ν΄μ£ΌμΈμ..."
Prompt Structure
기본 ꡬ쑰
## Role (μν )
[AIκ° μνν μν μ μ]
## Context (λ§₯λ½)
[λ°°κ²½ μ 보, μ μ½ μ‘°κ±΄]
## Task (μμ
)
[ꡬ체μ μΈ μμ² μ¬ν]
## Format (νμ)
[μνλ μΆλ ₯ νμ]
## Examples (μμ) - μ νμ
[μ
μΆλ ₯ μμ]
μ½λ κ΄λ ¨ ν둬ννΈ
## νκ²½
- μΈμ΄/νλ μμν¬: [λͺ
μ]
- λ²μ : [λͺ
μ]
- κ΄λ ¨ μμ‘΄μ±: [λͺ
μ]
## νμ¬ μν
[νμ¬ μ½λ λλ μν©]
## λͺ©ν
[λ¬μ±νλ €λ κ²°κ³Ό]
## μ μ½ μ‘°κ±΄
- [μ μ½ 1]
- [μ μ½ 2]
Anti-Patterns
νΌν΄μΌ ν ν¨ν΄
| μν°ν¨ν΄ | λ¬Έμ μ | κ°μ |
|---|---|---|
| λͺ¨νΈν μμ² | ν΄μμ μ¬μ§κ° λ§μ | ꡬ체μ 쑰건 λͺ μ |
| κ³Όλν μ 보 | ν΅μ¬μ΄ 묻ν | κ΄λ ¨ μ λ³΄λ§ ν¬ν¨ |
| λΆμ ν μ§μ | "~νμ§ λ§μΈμ" | "~ν΄μ£ΌμΈμ" λ‘ λ³κ²½ |
| μ΄λ¦° κ²°λ§ | λ²μκ° λ¬΄νν¨ | λ²μμ μ μ½ λͺ μ |
μμ
β "μ΄μν μ½λ ν¨ν΄ μ¬μ©νμ§ λ§μΈμ"
β
"ESLint κ·μΉμ λ°λ₯΄κ³ , ν¨μλ 20μ€ μ΄λ΄λ‘ μμ±ν΄μ£ΌμΈμ"
β "μ’μ μ±λ₯μΌλ‘ λ§λ€μ΄μ£ΌμΈμ"
β
"μλ΅ μκ° 100ms μ΄λ΄, λ©λͺ¨λ¦¬ μ¬μ© 512MB μ΄νλ‘ μ΅μ νν΄μ£ΌμΈμ"
System Prompt Design
ꡬ쑰
# Identity
[AIμ μ 체μ±κ³Ό λͺ©μ ]
# Capabilities
[ν μ μλ κ²]
# Constraints
[μ ν μ¬ν]
# Response Format
[μλ΅ νμ κ·μΉ]
# Examples
[λνμ μΈ μνΈμμ© μμ]
μμ: μ½λ 리뷰 λ΄
# Identity
λΉμ μ μ½λ νμ§ ν₯μμ λλ μλμ΄ κ°λ°μμ
λλ€.
# Capabilities
- μ½λ μ€νμΌ κ²ν
- λ²κ·Έ νμ§
- μ±λ₯ κ°μ μ μ
- 보μ μ·¨μ½μ μλ³
# Constraints
- κ°μΈμ μ€νμΌ κ°μ κΈμ§
- νλ‘μ νΈ μ»¨λ²€μ
μ°μ
- 건μ€μ νΌλλ°±λ§ μ 곡
# Response Format
## μμ½
[ν μ€ μμ½]
## μ£Όμ λ°κ²¬
- [μ΄μ 1]: [μ€λͺ
] β [μ μ]
## κΈμ μ μΈ‘λ©΄
- [μν μ ]
Iteration Strategies
1. μ΄κΈ° ν둬ννΈ μμ±
2. κ²°κ³Ό νκ°
3. λ¬Έμ μ μλ³:
- λΆμμ ν μλ΅ β λ ꡬ체μ μΈ μ§μ μΆκ°
- μλͺ»λ λ°©ν₯ β μ μ½ μ‘°κ±΄ κ°ν
- νμ λΆμΌμΉ β μμ μΆκ°
4. ν둬ννΈ μμ
5. λ°λ³΅