The fastest, cheapest, most efficient open-source agent harness. Run more than 500 agents on a $10 VPS.
Rust agent-orchestrationai-agentsai-assistantdesktopllmlocal-firstmcppersonal-aiprivacyrustsecond-braintauri
/**
* CoreTransport interface — all core-RPC transports implement this.
*
* Implementations:
* LocalTransport — local HTTP to the in-process core sidecar
* LanHttpTransport — HTTP to a LAN-accessible core URL
* TunnelTransport — socket.io E2E-encrypted relay
* CloudHttpTransport — HTTP to a user-configured cloud core URL
*/
export type TransportKind = 'local' | 'lan-http' | 'tunnel' | 'cloud-http';
export interface CoreTransport {
readonly kind: TransportKind;
/** Make a JSON-RPC call and return the result. */
/**
* `timeoutMs`, when given, replaces the transport's own default deadline for
* this one call. The RPC client forwards a caller's per-call budget here so a
* request that legitimately runs for minutes (a memory source sync, a
* coding-session import) is not cut off at the transport's default while the
* core keeps working on it.
*/
call<T>(
method: string,
params: unknown,
opts?: { signal?: AbortSignal; timeoutMs?: number }
): Promise<T>;
/** Stream a JSON-RPC method that produces sequential chunks. */
stream<T>(method: string, params: unknown, opts?: { signal?: AbortSignal }): AsyncIterable<T>;
/** Probe the transport with a ping. */
isHealthy(): Promise<boolean>;
/** Tear down the transport. */
close(): Promise<void>;
}