# PromptSmith

## Docs

- [createPromptBuilder()](https://mintlify.wiki/galfrevn/promptsmith/api/create-prompt-builder.md): Factory function to create a new SystemPromptBuilder instance for building type-safe AI agent prompts
- [Composition](https://mintlify.wiki/galfrevn/promptsmith/api/methods/composition.md): Extend and merge builders for reusable configurations
- [Constraints](https://mintlify.wiki/galfrevn/promptsmith/api/methods/constraints.md): Define behavioral rules and guidelines for your agent
- [Context & Examples](https://mintlify.wiki/galfrevn/promptsmith/api/methods/context-examples.md): Provide domain knowledge and behavioral examples
- [Identity & Capabilities](https://mintlify.wiki/galfrevn/promptsmith/api/methods/identity-capabilities.md): Define your agent's identity and core capabilities
- [Output & Tone](https://mintlify.wiki/galfrevn/promptsmith/api/methods/output-tone.md): Configure response formatting and communication style
- [Tools](https://mintlify.wiki/galfrevn/promptsmith/api/methods/tools.md): Register tools that your agent can use
- [Validation & Export](https://mintlify.wiki/galfrevn/promptsmith/api/methods/validation.md): Build, validate, and export your prompts
- [PromptTester](https://mintlify.wiki/galfrevn/promptsmith/api/prompt-tester.md): Testing framework for evaluating system prompts against test cases using LLM-as-judge
- [PromptValidator](https://mintlify.wiki/galfrevn/promptsmith/api/prompt-validator.md): Validation framework for checking builder configuration and providing helpful error messages
- [SystemPromptBuilder](https://mintlify.wiki/galfrevn/promptsmith/api/system-prompt-builder.md): Main builder class for constructing type-safe system prompts for AI agents
- [Constraint Types](https://mintlify.wiki/galfrevn/promptsmith/api/types/constraint-types.md): Types for defining behavioral constraints and guidelines for AI agents
- [Format Types](https://mintlify.wiki/galfrevn/promptsmith/api/types/format-types.md): Types for configuring prompt output format and AI SDK integration
- [Test Types](https://mintlify.wiki/galfrevn/promptsmith/api/types/test-types.md): Types for testing and evaluating AI agent prompts
- [ToolDefinition & ExecutableToolDefinition](https://mintlify.wiki/galfrevn/promptsmith/api/types/tool-definition.md): Type definitions for declaring tools available to AI agents
- [SystemPromptBuilder](https://mintlify.wiki/galfrevn/promptsmith/concepts/builder.md): Understanding the fluent API for constructing type-safe AI agent prompts
- [Behavioral Constraints](https://mintlify.wiki/galfrevn/promptsmith/concepts/constraints.md): Define rules and guidelines that govern AI agent behavior with different severity levels
- [Output Formats](https://mintlify.wiki/galfrevn/promptsmith/concepts/formats.md): Optimize token usage with markdown, TOON, and compact formats for system prompts
- [Tool Integration](https://mintlify.wiki/galfrevn/promptsmith/concepts/tools.md): How to define, validate, and document AI agent tools using Zod schemas
- [Code Assistant](https://mintlify.wiki/galfrevn/promptsmith/examples/code-assistant.md): Build a coding assistant with file operations and code analysis capabilities
- [Conditional Logic](https://mintlify.wiki/galfrevn/promptsmith/examples/conditional-logic.md): Learn how to use withToolIf and withConstraintIf to build dynamic prompts that adapt based on runtime conditions and configuration
- [Customer Support Agent](https://mintlify.wiki/galfrevn/promptsmith/examples/customer-support.md): Build a customer support agent with knowledge base search and order management
- [Data Analysis Agent](https://mintlify.wiki/galfrevn/promptsmith/examples/data-analysis.md): Build a data analysis agent with conditional logic and role-based access control
- [Dynamic Tools](https://mintlify.wiki/galfrevn/promptsmith/examples/dynamic-tools.md): Learn how to register tools at runtime, build dynamic tool collections, and create adaptive agents that respond to changing contexts
- [Error Handling](https://mintlify.wiki/galfrevn/promptsmith/examples/error-handling.md): Learn best practices for error handling in prompts, including validation, graceful degradation, and user-friendly error messages
- [Multi-Tool Agent](https://mintlify.wiki/galfrevn/promptsmith/examples/multi-tool-agent.md): Build agents with multiple tools using both AI SDK and Mastra frameworks
- [Examples Overview](https://mintlify.wiki/galfrevn/promptsmith/examples/overview.md): Complete examples showing how to build AI agents with PromptSmith
- [Prompt Composition](https://mintlify.wiki/galfrevn/promptsmith/examples/prompt-composition.md): Learn how to use extend() and merge() to create reusable prompt patterns and compose complex prompts from modular components
- [Building Prompts](https://mintlify.wiki/galfrevn/promptsmith/guides/building-prompts.md): Step-by-step guide to constructing effective AI agent prompts with PromptSmith
- [Prompt Composition](https://mintlify.wiki/galfrevn/promptsmith/guides/composition.md): Extend and merge builders to create reusable prompt components and specialized agents
- [Security Guardrails](https://mintlify.wiki/galfrevn/promptsmith/guides/security-guardrails.md): Protect your AI agents from prompt injection and implement security best practices
- [Testing Prompts](https://mintlify.wiki/galfrevn/promptsmith/guides/testing.md): Validate AI agent behavior with PromptTester and LLM-as-judge evaluation
- [Token Optimization](https://mintlify.wiki/galfrevn/promptsmith/guides/token-optimization.md): Reduce API costs by 30-60% with TOON format and optimization strategies
- [Tool Integration](https://mintlify.wiki/galfrevn/promptsmith/guides/tool-integration.md): Add executable tools to your AI agents with type-safe Zod schemas
- [Installation](https://mintlify.wiki/galfrevn/promptsmith/installation.md): Install PromptSmith and get started building type-safe AI agents
- [Other Framework Integrations](https://mintlify.wiki/galfrevn/promptsmith/integrations/frameworks.md): Use PromptSmith with any AI framework or custom implementation
- [Mastra Integration](https://mintlify.wiki/galfrevn/promptsmith/integrations/mastra.md): Integrate PromptSmith with Mastra framework for advanced agent workflows
- [Vercel AI SDK Integration](https://mintlify.wiki/galfrevn/promptsmith/integrations/vercel-ai-sdk.md): Use PromptSmith with Vercel AI SDK for type-safe agent development with toAiSdk()
- [Introduction to PromptSmith](https://mintlify.wiki/galfrevn/promptsmith/introduction.md): Type-safe system prompt builder for production AI agents with Vercel AI SDK and Mastra integration
- [Quick Start](https://mintlify.wiki/galfrevn/promptsmith/quickstart.md): Build your first AI agent with PromptSmith in 5 minutes
- [Accessibility Template](https://mintlify.wiki/galfrevn/promptsmith/templates/accessibility.md): Reusable accessibility template for creating inclusive, screen-reader friendly interactions
- [Coding Assistant Template](https://mintlify.wiki/galfrevn/promptsmith/templates/coding-assistant.md): Expert coding assistant optimized for code writing, debugging, and technical explanations
- [Customer Service Template](https://mintlify.wiki/galfrevn/promptsmith/templates/customer-service.md): Professional customer service assistant optimized for e-commerce and support scenarios
- [Data Analyst Template](https://mintlify.wiki/galfrevn/promptsmith/templates/data-analyst.md): Data analysis assistant optimized for insights, visualization, and statistical analysis
- [Multilingual Template](https://mintlify.wiki/galfrevn/promptsmith/templates/multilingual.md): Reusable multilingual template for adding multi-language support to any prompt
- [Templates Overview](https://mintlify.wiki/galfrevn/promptsmith/templates/overview.md): Pre-built templates for common AI agent use cases
- [Research Assistant Template](https://mintlify.wiki/galfrevn/promptsmith/templates/research-assistant.md): Academic research assistant optimized for literature review and scholarly analysis
- [Security Template](https://mintlify.wiki/galfrevn/promptsmith/templates/security.md): Reusable security template for adding guardrails and best practices to any prompt