ActionCallbacksLearn how to use post-action execution with the `run` keyword to chain actions and process results.
ActionDefinitionsThis recipe demonstrates how to define and use actions - the way Agent Script agents call external systems, APIs, and services.
ActionDescriptionOverridesMaster context-specific action descriptions to improve LLM action selection.
BeforeAfterReasoningLearn how to use reasoning lifecycle events with `before_reasoning` and `after_reasoning`.
ErrorHandlingThis recipe demonstrates comprehensive error handling and validation patterns for building safe, reliable agents.
HelloWorldThis recipe demonstrates the absolute minimum structure required for an Agent Script agent.
MultiTopicNavigationThis recipe demonstrates how to build agents with multiple topics and navigate between them.
ReasoningInstructionsThis recipe demonstrates procedural reasoning instructions - a powerful way to build dynamic, context-aware agents.
SystemInstructionOverridesLearn how to override system instructions per topic to change agent behavior and persona dynamically.
TemplateExpressionsThis recipe demonstrates how to use template expressions to create dynamic, personalized content in your agent.