Struct Agent 

pub struct Agent<T: LLMBackend = OpenAI> {
    pub world: Option<World>,
    pub entity: Option<Entity>,
    /* private fields */
}

Fields

world: Option<World>

entity: Option<Entity>

Implementations

impl<T: LLMBackend> Agent<T>

pub fn builder() -> AgentBuilder<T>

pub async fn dispatch_user_message(&mut self, text: &str) -> String

pub async fn run(&mut self) -> Result<AgentRunOutput>

Run the agent to completion and return the final AgentRunOutput, which contains the full hecs::World with all components.

Errors

Returns an error if the LLM API call fails and no plugin handles it.

pub async fn run_json<R>(&mut self) -> Result<R>

where R: DeserializeOwned + JsonSchema,

Run the agent to completion and extract a structured JSON response.

This is a convenience wrapper around run that performs an extra LLM call with a response schema after the agent loop finishes.

Errors

Returns an error if the agent loop fails or JSON deserialization fails.

Trait Implementations

impl<T: LLMBackend> Debug for Agent<T>

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl<T: Default + LLMBackend> Default for Agent<T>

fn default() -> Agent<T>

Returns the “default value” for a type.