Compute robust operating viability kernel

What it's for

Moves Gitrevio beyond risk scores into robust operational control: leaders can see which operating states remain governable under every modeled disruption, which actions preserve the safe region, and how much response time remains outside it.

What you give it

Inputs split into evidence read from your connected systems, calibration your team owns, and numerical controls that affect precision but never the result's meaning.

{
  "actions": [
    {
      "id": "stabilize"
    },
    {
      "id": "accelerate"
    }
  ],
  "initial_state_ids": [
    "healthy",
    "strained"
  ],
  "required_safe_steps": 4,
  "states": [
    {
      "id": "healthy",
      "safe": true
    },
    {
      "id": "strained",
      "safe": true
    },
    {
      "id": "failed",
      "safe": false
    }
  ],
  "transitions": [
    {
      "action_id": "stabilize",
      "possible_next_state_ids": [
        "healthy"
      ],
      "state_id": "healthy"
    },
    {
      "action_id": "accelerate",
      "possible_next_state_ids": [
        "healthy",
        "strained"
      ],
      "state_id": "healthy"
    },

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.

{
  "assumptions": [
    "States are decision-sufficient at the chosen cadence, safe labels encode hard governed constraints, and every feasible action is represented for every state.",
    "possible_next_state_ids are a conservative set containing every materially possible successor under that state-action pair; omitted successors can create false safety.",
    "The transition envelope and action feasibility remain stable for the period in which the invariant policy is used."
  ],
  "decision": "all_initial_states_robustly_viable",
  "detail_counts": {
    "state_diagnostics": 3,
    "viable_policy_states": 2
  },
  "executive_summary": {
    "fixed_point_iterations": 1,
    "infinite_viability_kernel_states": 2,
    "initial_states": 2,
    "safe_states": 2,
    "states": 3,
    "viability_fraction_of_safe_states": 1,
    "viable_initial_states": 2
  },
  "initial_state_assessment": [
    {
      "action_exposure": [],
      "guaranteed_safe_steps": null,
      "in_infinite_viability_kernel": true,
      "meets_required_safe_horizon": true,
      "removal_iteration": null,
      "robustly_safe_action_ids": [
        "stabilize",
        "accelerate"
      ],
      "safe_now": true,
      "state_id": "healthy"
    },
    {
      "action_exposure": [
        {
          "action_id": "accelerate",
          "immediate_unsafe_successors": [
            "failed"
          ],
          "successors_outside_infinite_kernel": [
            "failed"
          ]

How it works

Forecasting & survival — Estimate when something completes or fails, with censoring and unresolved work handled honestly rather than dropped.

1. 1 Start with every state satisfying the governed hard safety constraints and require one complete, nonempty set of materially possible successors for every state-action pair.
2. 2 Apply the robust predecessor operator repeatedly: retain a safe state only when at least one feasible action has every possible successor inside the previous safe set.
3. 3 Continue the monotone elimination to a fixed point; the result is the maximal robust controlled-invariant viability kernel, and actions whose full successor set remains inside it form the safe feedback policy.
4. 4 Record the elimination iteration for nonviable states to show finite guaranteed-safe steps, evaluate the governed initial states and required finite horizon, and expose unsafe or nonviable successors by action rather than inventing a least-bad safe claim.

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 18 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": "compute the maximal robust controlledinvariant set" }
  → finds "compute_robust_operating_viability_kernel"

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

gitrevio_capability_run
  { "capability_id": "compute_robust_operating_viability_kernel", "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 Measurement integrity →