Skip to main content

ComposablePolicy

@totemsdk/agent-policy

@totemsdk/agent-policy / ComposablePolicy

Class: ComposablePolicy

ComposablePolicy chains multiple PolicyMiddleware layers into a single evaluation pipeline. Layers are evaluated in registration order with short-circuit semantics: if any layer returns rejected, subsequent layers are skipped and the rejection is returned immediately.

ComposablePolicy also implements the legacy AgentPolicy interface (canAutoApprove / requiresUserApproval) so it can be used anywhere the old interface is expected (e.g. @totemsdk/omnia's executeIntent).

Middleware contract​

  • approved → continue to next layer
  • rejected → short-circuit, return rejection
  • requires_human → short-circuit, return requires_human

An empty middleware list approves all proposals (pass-through).

Implements​

Constructors​

Constructor​

new ComposablePolicy(layers): ComposablePolicy

Parameters​

layers​

PolicyMiddleware[]

Returns​

ComposablePolicy

Methods​

canAutoApprove()​

canAutoApprove(proposal): Promise<boolean>

Return true if the wallet should sign the intent without user interaction. Implementations typically check risk, amount thresholds, and known agents.

Parameters​

proposal​

AgentProposal

Returns​

Promise<boolean>

Implementation of​

AgentPolicy.canAutoApprove

commit()​

commit(operationId): Promise<void>

Commit a prior reservation after execution succeeds.

Parameters​

operationId​

string

Returns​

Promise<void>

Implementation of​

PolicyMiddleware.commit

evaluate()​

evaluate(proposal): Promise<PolicyEvalResult>

Evaluate a proposal. Called in sequence by ComposablePolicy.

Parameters​

proposal​

AgentProposal

Returns​

Promise<PolicyEvalResult>

Implementation of​

PolicyMiddleware.evaluate

release()​

release(operationId): Promise<void>

Release a prior reservation after execution fails or is cancelled.

Parameters​

operationId​

string

Returns​

Promise<void>

Implementation of​

PolicyMiddleware.release

requiresUserApproval()​

requiresUserApproval(proposal): Promise<boolean>

Return true if the wallet must show a user-approval UI before signing. Generally the complement of canAutoApprove, but may have independent logic (e.g. always require approval for settlements regardless of risk).

Parameters​

proposal​

AgentProposal

Returns​

Promise<boolean>

Implementation of​

AgentPolicy.requiresUserApproval

reserve()​

reserve(proposal): Promise<PolicyEvalResult>

Optional reservation lifecycle for stateful policies.

The execution boundary calls reserve before executing, then commit after a successful execution or release on any failure path. When implemented, quota is only consumed through reserve/commit — a read-only evaluate / canAutoApprove never touches the limits.

Parameters​

proposal​

AgentProposal

Returns​

Promise<PolicyEvalResult>

Implementation of​

PolicyMiddleware.reserve

reset()​

reset(): Promise<void>

Optional: reset internal state (useful in tests or at midnight rollover).

Returns​

Promise<void>

Implementation of​

PolicyMiddleware.reset