# Prompt Patterns

> A Pattern Language for Knowledge Engineering with Large Language Models. By Bilgin Ibryam.

This website is the companion catalog to the Prompt Patterns book. It captures reusable prompt design moves that are explicit, grounded, and composable — not magic phrases or model-specific tricks.

## Pages

- [Home](https://www.promptpatterns.dev): Overview of the Prompt Patterns project and what a prompt pattern is.
- [Prompting Principles](https://www.promptpatterns.dev/principles): Six core design principles behind effective prompting (OFP, ISP, MPP, QVP, ORP, IDP).
- [Patterns Catalog](https://www.promptpatterns.dev/patterns): Complete catalog of 27 prompt patterns organized by six categories.
- [Resources](https://www.promptpatterns.dev/resources): Curated books, courses, videos, papers, articles, and communities about prompt engineering.
- [Buy the Book](https://www.promptpatterns.dev/buy): Full book with deeper explanations, worked examples, and composition guidance.

## Pattern Categories

### Objective Framing
Patterns that set the model's worldview before any work begins.
- [Task Directive](https://www.promptpatterns.dev/task-directive): Defines the exact unit of work the model must perform.
- [Perspective Framing](https://www.promptpatterns.dev/perspective-framing): Sets the communicative perspective the model should adopt.
- [Judgment Criteria](https://www.promptpatterns.dev/judgment-criteria): Declares the criteria for evaluating evidence and handling uncertainty.
- [Constraint Scoping](https://www.promptpatterns.dev/constraint-scoping): Defines the boundaries within which the model is allowed to operate.
- [Example-Driven Specification](https://www.promptpatterns.dev/example-driven-specification): Specifies desired behavior through concrete examples rather than abstract rules.

### Information Supply
Patterns that address what information to provide and how to control relevance.
- [Context Loading](https://www.promptpatterns.dev/context-loading): Supplies external information the model lacks but must rely on.
- [Context Curation](https://www.promptpatterns.dev/context-curation): Specifies which provided information counts as relevant evidence.
- [Structural Segmentation](https://www.promptpatterns.dev/structural-segmentation): Separates the prompt into distinct regions by designated purpose.

### Method Prescription
Patterns that prescribe how the model should process the task.
- [Alternative Enumeration](https://www.promptpatterns.dev/alternative-enumeration): Requires surfacing multiple viable approaches before committing.
- [Hierarchical Decomposition](https://www.promptpatterns.dev/hierarchical-decomposition): Breaks a concept into constituent parts through containment.
- [Dependency Decomposition](https://www.promptpatterns.dev/dependency-decomposition): Solves problems by establishing prerequisite answers progressively.
- [Stepwise Decomposition](https://www.promptpatterns.dev/stepwise-decomposition): Breaks down a problem into ordered intermediate steps.
- [Multi-Path Reasoning](https://www.promptpatterns.dev/multi-path-reasoning): Evaluates multiple independent reasoning paths before committing.
- [Governing Abstraction](https://www.promptpatterns.dev/governing-abstraction): Derives a governing principle before solving the specific instance.

### Quality Validation
Patterns that intervene between output produced and output trusted.
- [Knowledge Externalization](https://www.promptpatterns.dev/knowledge-externalization): Requires the model to expose background knowledge before performing the task.
- [Deferred Commitment](https://www.promptpatterns.dev/deferred-commitment): States a decision early, then defers final commitment until evaluated.
- [Reflective Evaluation](https://www.promptpatterns.dev/reflective-evaluation): Requires the model to evaluate its own output before committing.
- [Claim Enumeration](https://www.promptpatterns.dev/claim-enumeration): Requires listing claims asserted as true for verification.

### Output Representation
Patterns that shape output into consumable, parseable forms.
- [Semantic Compression](https://www.promptpatterns.dev/semantic-compression): Compresses context preserving meaning while discarding surface wording.
- [Response Tail](https://www.promptpatterns.dev/response-tail): Reserves a consistent slot for secondary content in every response.
- [Answer Boundary](https://www.promptpatterns.dev/answer-boundary): Forces commitment to a final discrete answer after reasoning.
- [Output Template](https://www.promptpatterns.dev/output-template): Declares response structure in advance for predictable outputs.
- [Formal Representation](https://www.promptpatterns.dev/formal-representation): Constrains output to a rule-governed symbolic form.

### Interaction Design
Patterns that design how the interaction between user and model is structured.
- [Control Reversal](https://www.promptpatterns.dev/control-reversal): Transfers control of the conversation to the model.
- [Interpretation Grammar](https://www.promptpatterns.dev/interpretation-grammar): Defines explicit rules for how terms and notation are interpreted.
- [Meta Prompting](https://www.promptpatterns.dev/meta-prompting): Uses one prompt to generate or refine another prompt.
- [Specification Extraction](https://www.promptpatterns.dev/specification-extraction): Extracts the implicit specification from a finished artifact.