LLM Prompt Engineering Kit

$39

Prompt template library, chain-of-thought patterns, few-shot examples, prompt versioning system, and A/B testing framework.

📁 56 files

MarkdownYAMLPythonReactLLMOpenAI

📄 Product Preview

Try the interactive reader and demo tools below, or get the full product with all content unlocked.

📖 Interactive Reader (Free Preview) ⚙ Try Demo Tools 📦 Download Free Sample

📁 File Structure 56 files

llm-prompt-engineering-kit/ ├── LICENSE ├── README.md ├── free-sample.zip ├── guide/ │ ├── 01-prompt-patterns.md │ └── 02-versioning-workflow.md ├── guides/ │ ├── prompt-patterns.md │ └── versioning-workflow.md ├── index.html ├── prompts/ │ ├── agents/ │ │ ├── guardrail.yaml │ │ ├── planning_prompt.yaml │ │ ├── react_prompt.yaml │ │ ├── self_reflection.yaml │ │ ├── supervisor.yaml │ │ └── tool_selection.yaml │ ├── classification/ │ │ ├── intent.yaml │ │ ├── language.yaml │ │ ├── priority.yaml │ │ ├── sentiment.yaml │ │ ├── topic.yaml │ │ └── toxicity.yaml │ ├── coding/ │ │ ├── bug_fixing.yaml │ │ ├── code_generation.yaml │ │ ├── code_review.yaml │ │ ├── documentation.yaml │ │ ├── sql_generation.yaml │ │ └── test_generation.yaml │ ├── extraction/ │ │ ├── contact_extraction.yaml │ │ ├── data_parsing.yaml │ │ ├── entity_extraction.yaml │ │ ├── key_value_extraction.yaml │ │ ├── relationship_extraction.yaml │ │ └── table_extraction.yaml │ └── summarization/ │ ├── abstractive.yaml │ ├── bullet_points.yaml │ ├── changelog.yaml │ ├── executive_summary.yaml │ ├── meeting_minutes.yaml │ └── technical_docs.yaml ├── requirements.txt ├── src/ │ ├── __init__.py │ ├── __pycache__/ │ │ ├── __init__.cpython-312.pyc │ │ ├── ab_testing.cpython-312.pyc │ │ ├── chains.cpython-312.pyc │ │ ├── diff_tool.cpython-312.pyc │ │ ├── registry.cpython-312.pyc │ │ └── template_engine.cpython-312.pyc │ ├── ab_testing.py │ ├── chains.py │ ├── diff_tool.py │ ├── registry.py │ └── template_engine.py └── tests/ ├── __pycache__/ │ ├── test_registry.cpython-312.pyc │ └── test_template_engine.cpython-312.pyc ├── test_registry.py └── test_template_engine.py

📖 Documentation Preview README excerpt

LLM Prompt Engineering Kit

A complete prompt engineering toolkit: 30 battle-tested prompt templates, a

Python-based template engine with variable substitution and few-shot support,

a versioned prompt registry for tracking changes, an A/B testing harness for

comparing prompt variants, and a semantic diff tool for reviewing prompt edits.

What's Included

Prompt Template Library (30 templates)

Curated, production-ready prompt templates organized by domain:

Category	Templates	Use Cases
Extraction (6)	Entity extraction, data parsing, key-value, relationships, tables, contacts	Turning unstructured text into structured data
Classification (6)	Sentiment, topic, intent, priority, language, toxicity	Routing, triage, and content moderation
Summarization (6)	Executive summary, bullet points, abstractive, changelog, meeting minutes, tech docs	Condensing long content into actionable summaries
Coding (6)	Code review, code generation, bug fixing, SQL generation, test generation, documentation	Developer productivity and code quality
Agents (6)	ReAct, plan-and-execute, tool selection, supervisor, self-reflection, guardrails	Building AI agent systems

Every template includes:

System message with detailed instructions
User template with named variables
Few-shot examples (where applicable)
Token budget estimates
Model compatibility notes
Quality notes and gotchas

Python Toolkit

Module	Purpose
`template_engine.py`	Load YAML templates, render with variables, build message arrays
`registry.py`	Version prompts, track history, diff between versions, export/import
`chains.py`	Chain-of-thought and few-shot pattern builders
`ab_testing.py`	A/B test harness with metric collection and statistical comparison
`diff_tool.py`	Semantic diff between prompt versions (line-level and structural)

Guides

Prompt Patterns — Decision tree for choosing the right pattern (zero-shot, few-shot, CoT, ReAct, etc.)
Versioning Workflow — How to version, test, shadow-deploy, and roll back prompts

Quick Start

1. Load and Render a Template


from src.template_engine import TemplateEngine
from pathlib import Path

engine = TemplateEngine()
engine.load_directory(Path("prompts/"))

# Get a template and render it
template = engine.get("extraction.entity_extraction")
messages = template.build_messages(
    input_text="Acme Corp announced a $50M acquisition of DataCo in Austin, Texas."
)

*... continues with setup instructions, usage examples, and more.*

📄 Code Sample .py preview

src/ab_testing.py """ Prompt A/B testing harness with metric collection. When you have two versions of a prompt and need to know which performs better, this harness runs both against a test dataset and collects metrics (accuracy, latency, token usage, human ratings). The harness supports: - Side-by-side comparison of two prompt versions - Automated scoring with custom metric functions - Statistical significance testing (basic chi-squared) - Result export for further analysis - Integration with the PromptRegistry for version tracking """ from __future__ import annotations import json import logging import random import time from dataclasses import dataclass, field from typing import Any, Callable from src.template_engine import PromptTemplate logger = logging.getLogger(__name__) @dataclass class TestCase: """A single test input with optional expected output.""" input_variables: dict[str, Any] expected_output: str = "" metadata: dict[str, Any] = field(default_factory=dict) @dataclass class TestResult: """Result of running a single test case against a prompt variant.""" variant_name: str test_case: TestCase rendered_prompt: str llm_output: str latency_ms: float = 0.0 token_count: int = 0 score: float = 0.0 # From the scoring function metadata: dict[str, Any] = field(default_factory=dict) # ... 169 more lines ...

Buy Now — $39 Back to Products