Struct LoopGovernance 

pub struct LoopGovernance { /* private fields */ }

Governance over Engine::with_builtins() plus an optional host policy. LLM backends deliberately have no seam here: RUN LOOP from CAL is the deterministic pass — model-attached runs stay on the host surfaces where credentials live.

Implementations

impl LoopGovernance

pub fn new() -> Self

pub fn with_policy(policy: Policy) -> Self

Attach the same loop-policy.json the CLI takes (--policy) — a CAL-triggered run honors the same auto-apply grants, denies, and severity floors.

Trait Implementations

impl Default for LoopGovernance

fn default() -> Self

Returns the “default value” for a type.

impl GovernanceHost for LoopGovernance

fn run_loop( &self, store: &dyn CalStoreFacade, opts: &RunLoopOptions, ) -> Result<Value>

RUN LOOP — run the analysis pass. Returns the run result as JSON.

fn review( &self, store: &dyn CalStoreFacade, rec_hash: &str, decision: ReviewDecision, because: &str, ) -> Result<()>

APPROVE/REJECT <hash> BECAUSE "…".

fn apply( &self, store: &dyn CalStoreFacade, rec_hash: &str, because: &str, destructive_cap: bool, ) -> Result<bool>

APPLY <hash> BECAUSE "…". Returns whether the apply is rollbackable.

fn rollback( &self, store: &dyn CalStoreFacade, rec_hash: &str, because: &str, ) -> Result<()>

ROLLBACK <hash> BECAUSE "…".

fn describe(&self, store: &dyn CalStoreFacade, what: LoopInfo) -> Result<Value>

The DESCRIBE reads. Returns the info as JSON.