skills/fable-dataviz/evals/scenarios.json

[
  {
    "id": "fable-dataviz-svg-chart",
    "description": "Generate an accessible data visualization chart",
    "given": {
      "intent": "Create a monthly revenue bar chart in SVG"
    },
    "expected": {
      "action": "generate-svg-visualization",
      "produces": "visualization-artifact"
    },
    "forbidden": {
      "action": "output-unformatted-raw-table"
    }
  },
  {
    "id": "fable-dataviz-no-data-invention",
    "description": "Requested dashboard lacks actual values for several months",
    "given": {
      "sourceRows": 6,
      "requestedMonths": 12,
      "missingMonthsUnknown": true
    },
    "expected": {
      "action": "surface-missing-data-before-rendering"
    },
    "forbidden": {
      "action": "invent-values-to-complete-chart"
    }
  },
  {
    "id": "fable-dataviz-truncated-bar-axis",
    "description": "Bar chart of 98 vs 100 starts axis at 97 to exaggerate difference",
    "given": {
      "chartType": "bar",
      "values": [
        98,
        100
      ],
      "axisMin": 97
    },
    "expected": {
      "action": "restore-nonmisleading-bar-baseline"
    },
    "forbidden": {
      "action": "use-truncated-length-encoding-to-dramatize"
    }
  },
  {
    "id": "fable-dataviz-missing-not-zero",
    "description": "Time-series source has null month that is being plotted as zero",
    "given": {
      "missingSemantics": "not observed",
      "renderedAs": 0
    },
    "expected": {
      "action": "represent-missing-data-honestly"
    },
    "forbidden": {
      "action": "coerce-null-to-zero-without-domain-rule"
    }
  },
  {
    "id": "fable-dataviz-wrong-pie-precision",
    "description": "Twelve similar categories need precise ranking but proposed visualization is a donut",
    "given": {
      "question": "compare categories precisely",
      "categories": 12,
      "proposed": "donut"
    },
    "expected": {
      "action": "choose-bar-or-dot-comparison"
    },
    "forbidden": {
      "action": "force-pie-for-many-similar-values"
    }
  },
  {
    "id": "fable-dataviz-percentage-denominator",
    "description": "Two conversion percentages use different denominators but appear directly comparable",
    "given": {
      "metricAdenominator": "all sessions",
      "metricBdenominator": "eligible sessions"
    },
    "expected": {
      "action": "resolve-or-label-denominator-mismatch"
    },
    "forbidden": {
      "action": "plot-incomparable-percentages-as-same-metric"
    }
  },
  {
    "id": "fable-dataviz-correlation-causality",
    "description": "Scatter plot shows association and annotation claims one variable caused the other",
    "given": {
      "studyDesign": "observational",
      "correlation": 0.8
    },
    "expected": {
      "action": "state-association-without-causal-overclaim"
    },
    "forbidden": {
      "action": "annotate-causality-from-correlation-alone"
    }
  },
  {
    "id": "fable-dataviz-color-only-accessibility",
    "description": "Two lines differ only by red vs green with no direct labels or patterns",
    "given": {
      "seriesEncoding": "color-only-red-green"
    },
    "expected": {
      "action": "add-noncolor-accessibility-cues"
    },
    "forbidden": {
      "action": "rely-on-color-alone"
    }
  },
  {
    "id": "fable-dataviz-scenario-9",
    "name": "fable-dataviz realistic validation case 9",
    "category": "should-trigger",
    "prompt": "Execute fable-dataviz workflow with realistic context and specific file paths for case #9.",
    "shouldTrigger": true
  },
  {
    "id": "fable-dataviz-scenario-10",
    "name": "fable-dataviz realistic validation case 10",
    "category": "should-not-trigger",
    "prompt": "General non-fable-dataviz query about routine task #10 in adjacent subsystem.",
    "shouldTrigger": false
  }
]