Struct DecisionLoop 

pub struct DecisionLoop { /* private fields */ }

Deterministic coordinator for one evidence-to-command cycle.

Example

use termivar_scanner::{
    AdaptationLimits, BenefitScore, DecisionLoop, DecisionLoopConfig, ExperiencePolicy,
    PlanningContext, RiskScore,
};

let planning = PlanningContext::new(
    BenefitScore::from_percent(80)?,
    100,
    RiskScore::from_percent(40)?,
);
let config = DecisionLoopConfig::new(
    planning,
    AdaptationLimits::default(),
    ExperiencePolicy::default(),
    32,
)?;
let decision_loop = DecisionLoop::new(config);
assert!(decision_loop.planner().is_empty());

Implementations

impl DecisionLoop

pub fn new(config: DecisionLoopConfig) -> Self

Creates an empty coordinator with explicit limits.

pub fn with_components( config: DecisionLoopConfig, rules: RuleEngine, planner: AttackPlanner, verification: VerificationPipeline, adaptive: AdaptivePipeline, ) -> Self

Creates a coordinator from independently configured subsystems.

pub fn config(&self) -> DecisionLoopConfig

Returns the immutable configuration.

pub fn rules(&self) -> &RuleEngine

Returns the reasoning registry.

pub fn rules_mut(&mut self) -> &mut RuleEngine

Returns the mutable reasoning registry.

pub fn planner(&self) -> &AttackPlanner

Returns the attack planner.

pub fn planner_mut(&mut self) -> &mut AttackPlanner

Returns the mutable attack planner.

pub fn verification(&self) -> &VerificationPipeline

Returns the verification pipeline.

pub fn verification_mut(&mut self) -> &mut VerificationPipeline

Returns the mutable verification pipeline.

pub fn adaptive(&self) -> &AdaptivePipeline

Returns adaptive policy.

pub fn adaptive_mut(&mut self) -> &mut AdaptivePipeline

Returns mutable adaptive policy.

pub fn plan_next( &self, knowledge: &KnowledgeBase, experience: &ExperienceStore, session: &mut DecisionSession, ) -> Result<DecisionPlanningReport, DecisionLoopError>

Applies reasoning, utility planning, and target-scoped suppressions.

pub fn plan_next_with_suppressed_actions( &self, knowledge: &KnowledgeBase, experience: &ExperienceStore, session: &mut DecisionSession, host_suppressed_actions: &BTreeSet<String>, ) -> Result<DecisionPlanningReport, DecisionLoopError>

Applies reasoning and planning while honoring host policy suppressions.

Explicit suppressions are combined with experience and adaptive-session suppressions. They remain visible as policy exclusions in the returned planner audit record. A host that later submits passive or active evidence must use the matching suppression-aware submission method so adaptive execution is reauthorized against current policy.

pub fn submit_passive( &self, knowledge: &KnowledgeBase, experience: &mut ExperienceStore, session: &mut DecisionSession, ) -> Result<DecisionOutcomeReport, DecisionLoopError>

Evaluates evidence produced by the outstanding action.

pub fn submit_passive_with_suppressed_actions( &self, knowledge: &KnowledgeBase, experience: &mut ExperienceStore, session: &mut DecisionSession, host_suppressed_actions: &BTreeSet<String>, ) -> Result<DecisionOutcomeReport, DecisionLoopError>

Evaluates passive evidence under an explicit current host policy.

Adaptive directives that continue automated work require this explicit context, even when the set is empty. This prevents replay or a mixed planning/submission API from silently forgetting host authority.

pub fn submit_active( &self, knowledge: &KnowledgeBase, experience: &mut ExperienceStore, session: &mut DecisionSession, baseline: &KnowledgeSnapshot, after_probe: &KnowledgeSnapshot, ) -> Result<DecisionOutcomeReport, DecisionLoopError>

Evaluates evidence produced by an explicit active verification probe.

pub fn submit_active_with_suppressed_actions( &self, knowledge: &KnowledgeBase, experience: &mut ExperienceStore, session: &mut DecisionSession, baseline: &KnowledgeSnapshot, after_probe: &KnowledgeSnapshot, host_suppressed_actions: &BTreeSet<String>, ) -> Result<DecisionOutcomeReport, DecisionLoopError>

Evaluates active evidence under an explicit current host policy.

See Self::submit_passive_with_suppressed_actions for why adaptive execution requires an explicit, replay-time host authority context.

Trait Implementations

impl Clone for DecisionLoop

fn clone(&self) -> DecisionLoop

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for DecisionLoop

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

Formats the value using the given formatter.