← WebMCP Directory

Tool tree

gridfall-st-anne.xmichal-severin.chatgpt.site

AI & AgentsTools verified
Website
A shared city resilience simulator where people and agents diagnose outages, explore recovery futures and restore critical services together.
Explore capabilities
    get_city_stateAnswer

    Inspect the public GRIDFALL mission state, incident clues, infrastructure nodes, and recovery progress. The hidden solution is never exposed.

    View tool JSON
    {
      "name": "get_city_state",
      "kind": "answer",
      "impl": "imperative",
      "description": "Inspect the public GRIDFALL mission state, incident clues, infrastructure nodes, and recovery progress. The hidden solution is never exposed."
    }
    inspect_hospital_statusAnswer

    Inspect ICU reserve, protected supply, live recovery blockers, telemetry, access, shelter, police, and cooling-water pressure.

    View tool JSON
    {
      "name": "inspect_hospital_status",
      "kind": "answer",
      "impl": "imperative",
      "description": "Inspect ICU reserve, protected supply, live recovery blockers, telemetry, access, shelter, police, and cooling-water pressure."
    }
    list_grid_actionsAnswer

    List the operational decisions currently available, including time cost, benefit, and systemic risk. This does not reveal the correct sequence.

    View tool JSON
    {
      "name": "list_grid_actions",
      "kind": "answer",
      "impl": "imperative",
      "description": "List the operational decisions currently available, including time cost, benefit, and systemic risk. This does not reveal the correct sequence."
    }
    take_grid_actionAction

    Submit one grid operation. In Human Approval mode it becomes a proposal that only the human operator can authorize in the web UI; in Agent Autopilot test mode it executes immediately. Never claim a Human Approval proposal was executed.

    View tool JSON
    {
      "name": "take_grid_action",
      "kind": "act",
      "impl": "imperative",
      "description": "Submit one grid operation. In Human Approval mode it becomes a proposal that only the human operator can authorize in the web UI; in Agent Autopilot test mode it executes immediately. Never claim a Human Approval proposal was executed.",
      "inputSchema": {
        "type": "object",
        "properties": {
          "action": {
            "type": "string",
            "enum": [
              "isolate_tx4",
              "shed_metro",
              "close_southbank_tie",
              "close_central_bus",
              "open_relay_r7",
              "deploy_battery_icu",
              "isolate_hospital_feeder",
              "deploy_battery_data",
              "reboot_scada",
              "shed_waterworks",
              "reroute_bank_alarm",
              "shed_bank",
              "activate_shelter",
              "prioritize_ambulances",
              "isolate_central_bus",
              "isolate_clinic",
              "shed_data_center",
              "dispatch_police",
              "close_harbor_tie"
            ]
          }
        },
        "required": [
          "action"
        ],
        "additionalProperties": false
      }
    }
    start_random_incidentAction

    Start a 15-second pre-blackout countdown and draw one of 100 validated incident puzzles.

    View tool JSON
    {
      "name": "start_random_incident",
      "kind": "act",
      "impl": "imperative",
      "description": "Start a 15-second pre-blackout countdown and draw one of 100 validated incident puzzles."
    }
    reset_missionAction

    Return GRIDFALL to standby. The next start draws a different incident.

    View tool JSON
    {
      "name": "reset_mission",
      "kind": "act",
      "impl": "imperative",
      "description": "Return GRIDFALL to standby. The next start draws a different incident."
    }