Async-RBench: delayed callbacks and lifecycle correctness of agent workflows

Evidence record

Async-RBench: delayed callbacks and lifecycle correctness of agent workflows

Read the original source ↗

Claim-level citations (2)

• supportsAsync agent callbacks reduce correct task completion in a controlled benchmark: Async-RBench defines Event Response Contracts that separately score what a delayed agent response must change, preserve, avoid and re-verify; duplicate reports and subagent failures are the hardest evaluated categories.

  Abstract, Event Response Contract and Dynamic Replanning Score

• supportsAsync agent callbacks reduce correct task completion in a controlled benchmark: The authors evaluate 200 executable tasks under eight asynchronous disruption scenarios with nine agents, reporting average task success of 26.0 percent compared with 29.7 percent for the paired batched condition.

  Abstract; 200-task asynchronous evaluation; measured success and state outcomes