Operational Run Situations in StateM
At any point during execution, a StateM agent run occupies one of three distinct operational situations:
- Active: The agent is executing within a non-terminal state and progressing through phase obligations.
- Paused: Execution is halted in a non-terminal state because an external condition, dependency, or human approval is unresolved.
- Complete: Execution has reached a designated terminal state and satisfied all exit verification contracts.
Critically, StateM distinguishes between an agent being paused due to an unresolved condition and an agent completing its work. Reaching a pause in a non-terminal state is not treated as a successful completion, preventing temporary blockages from being misclassified as valid handoffs.
0
1
Tags
Prep Sessions
Long-Horizon Agent Reliability: Stateful Scaffolding and Runtime Verification @ University of Michigan - Ann Arbor
Ch.3 State Persistence and Continuous Optimization - Long-Horizon Agent Reliability: Stateful Scaffolding and Runtime Verification @ University of Michigan - Ann Arbor
Persistent Run State and Recoverable Execution Protocols - Long-Horizon Agent Reliability: Stateful Scaffolding and Runtime Verification @ University of Michigan - Ann Arbor
Learn After
In StateM, which criteria must be met for an agent run to occupy the Complete operational situation?
In StateM, an agent halted in a non-terminal state due to an unresolved human approval is classified as Complete.
According to StateM, what operational situation does an agent run occupy when it is executing within a non-terminal state and progressing through phase obligations?