Calculate recommendation conflict exposure

What it's for

Turns contradictory dashboards or AI answers into an executive financial question: how much loss is exposed, whether the current decision is economically defensible, and what reserve or arbitration is required.

What you give it

Every tool is a pure function: it computes from what you pass it and reads nothing on its own. Inputs split into evidence you supply from your connected systems, calibration your team owns, and numerical controls that affect precision but never the result's meaning.

{
  "actions": [
    {
      "id": "accelerate",
      "loss_scenarios": [
        0,
        100
      ]
    },
    {
      "id": "stage",
      "loss_scenarios": [
        60,
        20
      ]
    }
  ],
  "current_action_id": "stage",
  "current_conflict_risk_reserve": 100,
  "recommendations": [
    {
      "action_id": "accelerate",
      "calibrated_reliability": 0.8,
      "evidence_source_ids": [
        "delivery-history"
      ],
      "id": "delivery",
      "supported": true,
      "weight": 1
    },
    {
      "action_id": "stage",
      "calibrated_reliability": 0.9,
      "evidence_source_ids": [
        "finance-case"
      ],
      "id": "finance",
      "supported": true,
      "weight": 1
    },
    {
      "action_id": "stage",
      "calibrated_reliability": 0.8,
      "evidence_source_ids": [

What you get back

This is the actual output of running the example above — computed by the same function the platform calls, not an illustration.

{
  "action_diagnostics": [
    {
      "action_id": "stage",
      "expected_loss": 40,
      "expected_regret_vs_scenario_oracle": 30,
      "is_consensus_action": true,
      "is_current_action": true,
      "is_minimum_expected_loss_action": true,
      "lineage_adjusted_recommendation_support": 0.6667,
      "probability_not_scenario_optimal": 0.5,
      "tail_cvar_regret_vs_scenario_oracle": 60
    },
    {
      "action_id": "accelerate",
      "expected_loss": 50,
      "expected_regret_vs_scenario_oracle": 40,
      "is_consensus_action": false,
      "is_current_action": false,
      "is_minimum_expected_loss_action": false,
      "lineage_adjusted_recommendation_support": 0.3333,
      "probability_not_scenario_optimal": 0.5,
      "tail_cvar_regret_vs_scenario_oracle": 80
    }
  ],
  "configuration": {
    "maximum_acceptable_cvar_regret": null,
    "minimum_consensus_support": 0.6,
    "reserve_confidence_level": 0.9,
    "scenario_count": 2,
    "tail_probability": 0.5
  },
  "decision": "recommendation_conflict_exposure_supported",
  "failed_gates": [],
  "guardrails": [
    "Loss scenarios must share one currency, horizon, perimeter and coherent common-shock column order; independently sorted action losses destroy regret and tail dependence.",
    "Scenario-oracle regret and perfect-information value are upper-bound diagnostics, not claims that perfect foresight or the ex-post best action was available.",
    "Reliability and lineage weights require prospective local validation. Consensus, reserve and minimum expected loss do not authorize an action that fails safety, legal, fiduciary or causal-evidence review."
  ],
  "lineage_adjusted_action_support": {
    "accelerate": 0.3333,
    "stage": 0.6667
  },
  "method": "provenance_adjusted_recommendation_conflict_exposure_v1",

How it works

Decision analysis — Turn uncertainty, cost and risk appetite into a defensible choice, with the reasoning left inspectable.

1. 1 Freeze one decision's feasible actions, coherent scenario losses, current action and prospectively calibrated recommendations with explicit evidence lineage, reliability, support and weight.
2. 2 Allocate one vote per unique evidence source, identify supported consensus without vote stuffing, and calculate each action's expected loss, scenario-oracle regret, suboptimal probability and regret CVaR.
3. 3 Compare the current action with the minimum-expected-loss action, size the conflict-risk reserve, price supported-action loss dispersion and withhold support when consensus, reserve, tail or expected-loss gates fail.

Before you trust it

Every tool in the catalog ships with the conditions under which its answer is meaningful — and the conditions under which it should abstain instead of guessing.

Calibrating it to your org

Calibration workflow

1. 1 Define the management decision, target outcome, aggregate unit, privacy boundary, cadence, and prediction/intervention horizon for this organization.
2. 2 Build a tenant-scoped historical cohort using only information available before each prediction or decision; preserve zero periods, censoring, assignment probabilities, and unresolved outcomes when the method requires them.
3. 3 Estimate statistical parameters on training history, but obtain costs, utilities, risk tolerance, practical-effect thresholds, capacity, and policy constraints from accountable decision owners.
4. 4 Validate on later time windows or held-out aggregate units at the deployment grain, against a simple baseline and the function-specific validation strategy.
5. 5 Deploy only if the returned decision clears evidence, overlap, calibration, robustness, and guardrail checks; warning, unsupported, schema-gap, and fallback decisions are abstentions.
6. 6 Monitor realized outcomes, data drift, coverage, and decision regret; recalibrate at a governed cadence or after a detected regime/definition change, never merely because a stakeholder dislikes the result.

Call it from your AI

You don't wire up 388 tools in your MCP client. The GitRevio MCP server exposes 20 tools, three of which let an agent search the catalog, read a tool's schema, and run it — so the assistant finds this one on its own.

gitrevio_capabilities_search
  { "q": "price the expected reservequantile and tailcvar" }
  → finds "calculate_recommendation_conflict_exposure"

gitrevio_capability_describe
  { "capability_id": "calculate_recommendation_conflict_exposure" }
  → returns the input schema and agent guidance shown on this page

gitrevio_capability_run
  { "capability_id": "calculate_recommendation_conflict_exposure", "arguments": { ... } }
  → returns the result shown above

Works in Claude Desktop, Claude Code, Cursor, Cline, Continue.dev, Goose and Aider. See the MCP server.

Related tools

See every tool in Analytics assurance & orchestration →