Skip to main content
The runnable examples include single-step and two-step planning declarations. The required pieces are:
  1. action_path, also included in transition input_paths.
  2. Observed transitions containing the actual action and measured result.
  3. Planning actions, or a supported dynamic actions_path.
  4. state_bindings mapping predicted targets into the next projected state.
  5. At least an appropriate goal/objective for the task.
  6. futures.enabled: true and planning enabled: true, probabilistic: true.
For example, the binding from values.next_x to values.x means the predicted next value becomes the current value at the next search step. Without the correct bindings, increasing horizon does not create a meaningful multi-step problem. Query with the current state and request transition_plan. For a planner-led decision, leave the action unset; the planner assigns each candidate action during evaluation. An explicitly action-conditioned forecast can take precedence for the stored forecast receipt, so do not forecast one action and report a different action as its execution.

Execute and replan

Read transition_plan.status. When it is planned:
  1. Take transition_plan.selected_action.
  2. Execute only that first action.
  3. Report the real result and that decision’s ID.
  4. Query again using the new observed state.
action_sequence is the candidate sequence explored by the model, not permission to execute all its actions without observing the environment again. This is bounded action-sequence planning with replanning, not a full contingent policy tree for every possible future observation. selection.selected_policy is a separate output from the contextual policy selector. Probabilistic planning does not silently replace it. An application using this planner must deliberately consume transition_plan.selected_action.

Inspect the plan

Inspect selected_action, action_sequence, expected_utility, lower_tail_utility, goal_probability, failure_probability, failure_probability_upper, unresolved_probability, probability_mass, trajectories, candidate_sequences and rejected_sequences. The selected action’s first-step distribution is available as transition_plan.futures. This is especially useful when the ordinary forecast abstains because no action was supplied in the query. Constraint-failing branches and unsupported probability mass remain visible. They are not silently dropped to make the remaining plan look safer. Probability estimates describe the learned model; they are not certified safety bounds.