Struct ScriptedToolExecutor 

pub struct ScriptedToolExecutor { /* private fields */ }

In-memory scripted tool executor fixture for SDK conformance tests. Use it to script deterministic behavior in memory; any transcript or endpoint mutation is documented on the method that performs it.

Implementations

impl ScriptedToolExecutor

pub fn new(executor_ref: ExecutorRef, output: ToolExecutionOutput) -> Self

Creates a new testing::tool value with explicit caller-provided inputs. This constructor is data-only and performs no I/O or external side effects.

Panics

Panics if constructor invariants fail, such as invalid identifier text or constructor-specific bounds. Use a fallible constructor such as try_new when one is available for untrusted input.

pub fn calls(&self) -> Vec<ToolExecutionRequest>

Operates on in-memory or journal-derived testing::tool state for diagnostics and repair evidence. It does not create a second run loop or product workflow owner.

pub fn call_count(&self) -> usize

Returns the call count currently held by this value. This reads deterministic in-memory test state and performs no external I/O.

Trait Implementations

impl Clone for ScriptedToolExecutor

fn clone(&self) -> ScriptedToolExecutor

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl ToolExecutor for ScriptedToolExecutor

fn executor_ref(&self) -> &ExecutorRef

Returns executor ref for the current value. This is a read-only or data-construction helper unless the method body explicitly calls a port or store.

fn execute( &self, request: &ToolExecutionRequest, ) -> Result<ToolExecutionOutput, AgentError>

Executes one policy-approved tool request through this executor. Implementations may run host code or external adapters, but the runtime owns intent/result journaling and approval checks around this call.