RefineAct
A runtime verification framework for checking an LLM agent's proposed actions before execution.
RefineAct studies LLM-agent reliability at the boundary between proposed action and executed action. It inserts a runtime verification layer that checks whether a candidate action is consistent with the task intent before the action is allowed to affect the environment.
Standard agent frameworks can execute a tool action as soon as the model produces it. If the action is unsafe or inconsistent with the task, the error may become visible only after the environment has already changed.
RefineAct inserts a verification layer between action generation and execution. Candidate actions are checked against a task-intent specification, and rejected actions are regenerated before they can run.
The work is reported in an ASE 2026 paper, and the public lightweight implementation is linked as aligned while the full paper artifact is prepared for release.
Checking actions at runtime provides an auditable control point for agent behavior, complementing prompt design and model training with an execution-time reliability mechanism.
aligned, a lightweight, non-formal implementation, is available now. The full code from the paper will be released shortly.
The formal approach and complete evaluation are in the paper (Batole et al., 2026).