pub struct LongRunningExecutionPolicy {
pub max_discovered_steps: Option<u32>,
pub max_discovery_depth: Option<u32>,
pub cost_alert_threshold_usd: Option<f64>,
pub idle_timeout_seconds: Option<u64>,
pub max_same_step_repetitions: Option<u32>,
}Expand description
Policy configuration for long-running agentic executions
These controls prevent runaway costs, infinite discovery loops, and idle executions from consuming resources in the Agentic DAG model.
Fields§
§max_discovered_steps: Option<u32>Maximum number of dynamically discovered steps before intervention (Steps with StepSource::Discovered)
max_discovery_depth: Option<u32>Maximum depth of discovery chains (prevents infinite discovery) e.g., agent discovers step A, which discovers step B, which discovers C…
cost_alert_threshold_usd: Option<f64>Alert threshold for cumulative cost in USD When exceeded, execution pauses for approval
idle_timeout_seconds: Option<u64>Maximum time without activity before idle timeout (seconds)
max_same_step_repetitions: Option<u32>Maximum repetitions of same methodology before loop detection (default: 3)
Implementations§
Source§impl LongRunningExecutionPolicy
impl LongRunningExecutionPolicy
Sourcepub fn standard() -> Self
pub fn standard() -> Self
Standard preset - balanced limits for typical long-running executions
- Max duration: ~30 minutes (via idle timeout)
- Discovered steps: 50
- Discovery depth: 5
- Cost alert: $5.00 USD
Sourcepub fn extended() -> Self
pub fn extended() -> Self
Extended preset - higher limits for complex, supervised workflows
- Max duration: ~4 hours
- Discovered steps: 300
- Discovery depth: 10
- Cost alert: $50.00 USD
Trait Implementations§
Source§impl Clone for LongRunningExecutionPolicy
impl Clone for LongRunningExecutionPolicy
Source§fn clone(&self) -> LongRunningExecutionPolicy
fn clone(&self) -> LongRunningExecutionPolicy
1.0.0 · Source§fn clone_from(&mut self, source: &Self)
fn clone_from(&mut self, source: &Self)
source. Read more