---

name: online-resource-scheduling description: Design deterministic online scheduling policies from current observations. Use when assigning arriving work to limited resources without seeing future requests.

Online Resource Scheduling

Use this skill to build online schedulers that make deterministic decisions from the current observation only.

Core Workflow

Convert each observation into a temporary state, rank pending work, score feasible actions by weighted marginal cost, update the temporary state immediately, then replay the final action list before returning it.

actions = []
temporary_state = copy_resources(observation)

for item in ranked_pending_items(observation):
  candidates = enumerate_feasible_actions(item, temporary_state)
  if not candidates:
    actions.append(defer_or_reject(item))
    continue

  scored = []
  for action in candidates:
    deltas = estimate_objective_deltas(action, temporary_state)
    score = sum(weights[k] * deltas[k] for k in deltas)
    scored.append((score, stable_tie_break(action), action))

  chosen = min(scored)[-1]
  actions.append(chosen)
  apply(chosen, temporary_state)

validate(actions, observation)
return actions

Weighted Marginal Scoring

When a task provides objective weights, use them to compare feasible actions. Avoid fixed rules such as "always first-fit", "always minimize fragmentation", or "always use the tightest slot". Those can be wrong when another objective component has a larger weighted effect.

Suggested generic workflow:

1. Read visible objective weights.
2. For each pending item, enumerate feasible actions.
3. For each feasible action, estimate the change in each objective component.
4. Compute weighted_marginal_score.
5. Choose the feasible action with the lowest score.
6. Apply the action to temporary state before scoring later actions.

weighted_marginal_score =
  weight_1 * delta_component_1
+ weight_2 * delta_component_2
+ weight_3 * delta_component_3
+ ...
+ deterministic_tie_break

Feasibility remains a hard filter. Only score feasible actions. Useful components might include resource activation cost, residual-capacity cost, waiting or lateness cost, rejection or unserved-work cost, and fragmentation or stranded-capacity cost.

Unrelated Example

In delivery planning, the shortest route is not always best. Suppose route distance has weight 1, but opening a new vehicle has weight 100. Sending a package on an already-open vehicle with 5 extra miles may be better than opening a new vehicle with only 1 extra mile:

weighted score =
  distance_weight * extra_distance
+ vehicle_weight * new_vehicle_used

The correct decision compares the weighted score, not distance alone.

Practical Guidance

• Use only information present in the current observation.
• Prefer deterministic tie-breaking so repeated runs are reproducible.
• If no feasible action exists, defer or reject rather than guessing.
• Validate the complete action list, not just each action in isolation.