← WebMCP Directory

Tool tree

mhtorregrosa.github.io

AI & AgentsTools verified
Website
Switchboard Zero is a deterministic human-agent crisis control room powered by WebMCP.
Explore capabilities
    inspect_gridAction

    Read the complete current Switchboard Zero simulation state, including damaged lines, district service, available generation, human-locked districts, risk tolerance and pending approvals. Use this before proposing or applying a recovery plan.

    View tool JSON
    {
      "name": "inspect_grid",
      "kind": "act",
      "impl": "imperative",
      "description": "Read the complete current Switchboard Zero simulation state, including damaged lines, district service, available generation, human-locked districts, risk tolerance and pending approvals. Use this before proposing or applying a recovery plan.",
      "inputSchema": {
        "type": "object",
        "properties": {}
      },
      "page": "/SwitchboardZero/"
    }
    simulate_recovery_planAction

    Build a deterministic recovery sequence without applying grid switches. Critical-first avoids all service trade-offs, balanced maximizes useful recovery with one explicit human decision, and maximum-coverage accepts elevated relay risk. The visible control room updates to show the proposed plan.

    View tool JSON
    {
      "name": "simulate_recovery_plan",
      "kind": "act",
      "impl": "imperative",
      "description": "Build a deterministic recovery sequence without applying grid switches. Critical-first avoids all service trade-offs, balanced maximizes useful recovery with one explicit human decision, and maximum-coverage accepts elevated relay risk. The visible control room updates to show the proposed plan.",
      "inputSchema": {
        "type": "object",
        "properties": {
          "strategy": {
            "type": "string",
            "enum": [
              "critical-first",
              "balanced",
              "maximum-coverage"
            ],
            "description": "Recovery objective to simulate."
          }
        },
        "required": [
          "strategy"
        ]
      },
      "page": "/SwitchboardZero/"
    }
    apply_authorized_actionsAction

    Execute recovery steps from the most recently simulated plan in order. Stop at the first step outside the current human-defined authority boundary. A later step is never executed past a withheld dependency.

    View tool JSON
    {
      "name": "apply_authorized_actions",
      "kind": "act",
      "impl": "imperative",
      "description": "Execute recovery steps from the most recently simulated plan in order. Stop at the first step outside the current human-defined authority boundary. A later step is never executed past a withheld dependency.",
      "inputSchema": {
        "type": "object",
        "properties": {
          "planId": {
            "type": "string",
            "description": "The exact plan ID returned by simulate_recovery_plan."
          }
        },
        "required": [
          "planId"
        ]
      },
      "page": "/SwitchboardZero/"
    }
    request_human_decisionAction

    Place the action that was actually withheld by apply_authorized_actions into the visible human decision queue. The action must belong to the latest plan. This tool explains the trade-off but never approves or performs it.

    View tool JSON
    {
      "name": "request_human_decision",
      "kind": "act",
      "impl": "imperative",
      "description": "Place the action that was actually withheld by apply_authorized_actions into the visible human decision queue. The action must belong to the latest plan. This tool explains the trade-off but never approves or performs it.",
      "inputSchema": {
        "type": "object",
        "properties": {
          "actionId": {
            "type": "string",
            "enum": [
              "transfer_transit_feed",
              "bypass_solar_interlock"
            ],
            "description": "Sensitive action returned as blocked by apply_authorized_actions."
          },
          "reason": {
            "type": "string",
            "minLength": 12,
            "maxLength": 280,
            "description": "Plain-language reason that states who gains, who loses and for how long or at what risk."
          }
        },
        "required": [
          "actionId",
          "reason"
        ]
      },
      "page": "/SwitchboardZero/"
    }
    simulate_delay_impactAction

    Run a non-destructive counterfactual showing what would happen if the operator waited between one and twelve minutes. This never advances the live exercise clock or changes the grid.

    View tool JSON
    {
      "name": "simulate_delay_impact",
      "kind": "act",
      "impl": "imperative",
      "description": "Run a non-destructive counterfactual showing what would happen if the operator waited between one and twelve minutes. This never advances the live exercise clock or changes the grid.",
      "inputSchema": {
        "type": "object",
        "properties": {
          "minutes": {
            "type": "integer",
            "minimum": 1,
            "maximum": 12
          }
        },
        "required": [
          "minutes"
        ]
      },
      "page": "/SwitchboardZero/"
    }