savings_methodology
Classify a declared savings mechanism as hard or soft against the buyer's own definitions, and say what evidence a case built on it is held to. Route here when the question is what kind of saving is being claimed: the kind that shows up in adjudicated paid claims, or the kind that exists in a model. Declare `mechanism`, one of unit_cost_reduction, utilisation_reduction, demand_redirection, avoidance_of_future_cost, overpayment_recovery and adherence_improvement. The vocabulary is closed and no prose is interpreted anywhere in this tool: a mechanism outside the vocabulary is refused naming it rather than guessed at, because the classification is what decides which evidence class a case is held to. The mapping is this world's own published definition of hard and soft, and the definition text is served with the answer so the classification can be checked against its source. `hello_world` with skill "hammer-savings-classification" returns those definitions in full, out of this same build. The payload carries no number anywhere. It says what declaring the mechanism commits the caller to, whether it is hard or soft, why that classification follows, and what evidence would support it. Whether the mechanism delivers anything is settled by that evidence class, which this world does not hold. Hard and soft are classified one mechanism at a time and never summed. A claims-evidenced reduction added to a modelled counterfactual is one number with two different evidence classes behind it, and a reader of the total cannot tell which dollars were observed and which were hoped. The same rule keeps deferred_versus_avoided from adding a moved cost to a prevented one.
Use this form for new arguments. Follow concrete links for known values. Fetch-only readers may be unable to submit new URLs.
Exact input contract
{
"additionalProperties": false,
"properties": {
"mechanism": {
"description": "How the saving is claimed to work, named from the closed vocabulary. The caller declares it; this tool reads no sentences.",
"enum": [
"unit_cost_reduction",
"utilisation_reduction",
"demand_redirection",
"avoidance_of_future_cost",
"overpayment_recovery",
"adherence_improvement"
],
"type": "string"
}
},
"required": [
"mechanism"
],
"type": "object"
}