Struct AgentHarness 

pub struct AgentHarness<'a, A: Agent> { /* private fields */ }

An internal adapter turning an Agent into a Harness<World = (), Setup = ()> so the easy run_suite path reuses the exact same panic-isolated runner as the full Harness path.

A deliberate explicit struct rather than a blanket impl<T: Agent> Harness for T: a blanket impl would conflict (coherence) with a host’s own Harness impl (e.g. AetherCore’s AetherHarness), so the adapter keeps the two trait families independent. The world is () (built once per case, inert) and the agent’s run failure is mapped onto Harness::run’s anyhow::Result.

Implementations

impl<'a, A: Agent> AgentHarness<'a, A>

pub fn new(agent: &'a A) -> Self

Wrap an agent reference as a Harness.

Trait Implementations

impl<'a, A: Debug + Agent> Debug for AgentHarness<'a, A>

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

Formats the value using the given formatter.

impl<A: Agent> Harness for AgentHarness<'_, A>

type World = ()

The mutable world a case runs against (e.g. an executor over a bare game world). Built by setup, mutated by run, then scored by a Scorer whose Scorer::World must match this type.

type Setup = ()

The host’s per-case setup description (the input that builds a World). Matches EvalCase::setup.

fn setup(&self, _setup: &())

Build a fresh world for one case from its setup. Called exactly once per case, immediately before run. A panic here fails only this case.

fn run(&self, instruction: &str, _world: &mut ()) -> Result<RunArtifacts>

Run instruction against world, mutating it, and return what the run produced (tool calls, final text, tokens, transcript). Wall-clock latency is timed by the runner around this call.