skills.rest
SanketAdlak
SanketAdlakCOMMUNITY
Explore all skills by this author →
Website𝕏@devsanket14GitHub

prompt-engineering-patterns

Design, test, and optimize LLM prompts using few-shot, chain-of-thought, and structured output patterns.

Updated Apr 23, 2026
Search Tags:#prompt engineering#pydantic#prompt optimization#chain-of-thought#llm#few-shot learning#structured output
One-click install
npx skills add https://github.com/SanketAdlak/PDMProjectDesign --skill prompt-engineering-patterns-sanketadlak
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: prompt-engineering-patterns
Source: https://github.com/SanketAdlak/PDMProjectDesign/tree/main/.agents/skills/prompt-engineering-patterns
Command: npx skills add https://github.com/SanketAdlak/PDMProjectDesign --skill prompt-engineering-patterns-sanketadlak

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires numpy, and includes scripts (resource) and references (resource) and assets (resource) components.

What problem does it solve? LLM prompts often produce inconsistent, unparseable, or low-quality outputs in production. This Skill provides proven patterns and tooling to design, test, and iteratively optimize prompts for reliability and performance. ## Core Features & Use Cases - Prompting Patterns: Implement few-shot learning with dynamic example selection, chain-of-thought reasoning with self-consistency, and structured JSON outputs enforced with Pydantic schemas. - Prompt Optimization: Run A/B tests and iterative refinement with the included optimization script, tracking accuracy, latency, token usage, and success rate metrics. - Template Systems: Build reusable prompt templates with variable interpolation, conditional sections, and role-based system prompts. - Use Case: You are building a sentiment analysis feature and outputs are inconsistent. Use this Skill to apply structured output with a Pydantic schema, add few-shot examples from the assets library, and A/B test prompt variations until accuracy exceeds your target. ## Quick Start Ask the AI to convert your existing prompt into a structured-output prompt with a Pydantic schema and then optimize it against a small test suite.

Frequently Asked Questions about prompt-engineering-patterns

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I get structured JSON output from an LLM?▼

Define a Pydantic schema describing the expected fields, then instruct the model to respond in JSON matching that schema and parse the response into the model. With LangChain you can use with_structured_output to enforce the schema automatically.

How do I improve LLM accuracy with few-shot prompting?▼

Select 2-5 examples closely matching your task using semantic similarity retrieval with embeddings, keep formatting consistent across examples, and stay within your token budget. Include edge-case examples to handle boundary inputs.

When should I use chain-of-thought prompting?▼

Use chain-of-thought for math, logical reasoning, multi-step planning, and debugging tasks where step-by-step reasoning improves accuracy. Skip it for simple factual lookups, creative writing, or latency-sensitive applications.

How do I A/B test two prompt variations?▼

Run both prompts against the same test suite and compare accuracy, latency, and token metrics. The included optimize-prompt.py script automates this with parallel test execution and statistical comparison of results.

Why does my LLM output fail JSON parsing?▼

Parsing fails when the model adds extra prose around the JSON or omits required fields. Fix it by adding explicit format instructions, showing a schema example in the prompt, and adding a fallback path that retries with a simpler prompt on validation errors.

Related Skills

Explore More Software Engineering Skills→
voidborne-dvoidborne-d

hamel-husain

Build production-grade eval sets with human-validated LLM-as-judge workflows.

Community
Advanced
giuseppe-trisciuogliogiuseppe-trisciuoglio
322

prompt-engineering

Engineer few-shot, chain-of-thought, and system prompts for LLM pipelines.

Community
Advanced
EngineerWithAIEngineerWithAI
4

rag-implementation

Integrate vector databases and embedding models to build RAG systems for LLM applications.

Official
Advanced
lebsrallebsral
11

ai-fine-tuning

Fine-tune models on labeled datasets using BootstrapFinetune workflows.

Community
Advanced