fix: post-#21 merge hygiene (schema duplicate + blocks_from_analysis)

.claude-plugin/plugin.json

@@ -1,6 +1,6 @@
 {
   "name": "data-toolkit",
-  "version": "0.8.1",
+  "version": "0.8.2",
   "description": "Extract, tidy, reconcile, analyse, visualise and convert messy data into clean, validated, audit-ready tables and insight briefs \u2014 computed by a deterministic engine that runs on your machine, with no third-party services. From Phronesis Applied.",
   "author": {
     "name": "Phronesis Applied",

CHANGELOG.md

@@ -1,5 +1,19 @@
 # Changelog
 
+## 0.8.2 — 2026-07-19
+
+**Post-#21 merge hygiene** — two leftovers from landing the analyse plan-surface
+PR after the viz/filter work:
+
+- **`analysis-plan.schema.json`**: drop the duplicate `allOf` rule that required
+  `column` for `numeric_summary` / `outliers_iqr` / `currency_mix` a second time
+  (the #21 rule that also covers `distribution` remains).
+- **`blocks_from_analysis`**: render `histogram`, `stacked_bar`, and
+  `scatter_chart`. `suggest_blocks_from_analysis` already proposed the first two
+  for `distribution` / `pivot`; the convenience helper raised
+  `unsupported block type` instead of drawing them. Agent runtime was unaffected
+  (`suggest` → `_viz_block`).
+
 ## 0.8.1 — 2026-07-19
 
 **Follow-ups to the viz/filter PR (#25)** — closing the plan-surface and handoff

schemas/analysis-plan.schema.json

@@ -1 +1 @@
-{"$schema":"https://json-schema.org/draft/2020-12/schema","$id":"https://data-toolkit.local/schemas/analysis-plan.schema.json","title":"Data Analyse operations","type":"array","items":{"type":"object","required":["op"],"properties":{"op":{"enum":["numeric_summary","outliers_iqr","breakdown","period_series","ageing","currency_mix","concentration","pivot","distribution","trend","percentile","cohort","correlation_matrix","rolling","gini","seasonality","join_on","compare_series","filter_rows"]},"name":{"type":"string"},"column":{"type":"string","minLength":1},"by":{"type":"string","minLength":1},"value":{"type":"string"},"top":{"type":"integer","minimum":1},"top_n":{"type":"integer","minimum":1},"k":{"type":["number","string"]},"cap":{"type":"integer","minimum":1},"date_col":{"type":"string","minLength":1},"grain":{"enum":["month","quarter","year"]},"dayfirst":{"type":"boolean"},"as_of":{"type":"string","minLength":1},"buckets":{"type":"array","minItems":1,"items":{"type":"integer","minimum":0}},"rows_col":{"type":"string","minLength":1},"cols_col":{"type":"string","minLength":1},"aggfunc":{"enum":["sum","count","mean"]},"q":{"oneOf":[{"type":"number","minimum":0,"maximum":1},{"type":"array","minItems":1,"items":{"type":"number","minimum":0,"maximum":1}}]},"id_col":{"type":"string","minLength":1},"columns":{"type":"array","minItems":2,"items":{"type":"string","minLength":1}},"window":{"type":"integer","minimum":1},"func":{"enum":["mean","sum","median"]},"on":{"oneOf":[{"type":"string","minLength":1},{"type":"array","minItems":1,"items":{"type":"string","minLength":1}}]},"how":{"enum":["inner","left"]},"a_value":{"type":"string","minLength":1},"b_value":{"type":"string","minLength":1},"a_label":{"type":"string","minLength":1},"b_label":{"type":"string","minLength":1},"left":{"type":"object","required":["date_col"],"properties":{"date_col":{"type":"string","minLength":1},"value":{"type":"string"},"dayfirst":{"type":"boolean"}},"additionalProperties":false},"right":{"type":"object","required":["date_col"],"properties":{"date_col":{"type":"string","minLength":1},"value":{"type":"string"},"dayfirst":{"type":"boolean"}},"additionalProperties":false},"filters":{"type":"array","minItems":1,"items":{"type":"object","required":["op"],"properties":{"col":{"type":"string","minLength":1},"column":{"type":"string","minLength":1},"op":{"type":"string"},"value":{},"values":{"type":"array"},"lo":{},"hi":{}},"additionalProperties":false}}},"allOf":[{"if":{"properties":{"op":{"enum":["numeric_summary","outliers_iqr","currency_mix","distribution"]}},"required":["op"]},"then":{"required":["column"]}},{"if":{"properties":{"op":{"const":"breakdown"}},"required":["op"]},"then":{"required":["by"]}},{"if":{"properties":{"op":{"const":"period_series"}},"required":["op"]},"then":{"required":["date_col"]}},{"if":{"properties":{"op":{"const":"ageing"}},"required":["op"]},"then":{"required":["date_col","as_of"]}},{"if":{"properties":{"op":{"enum":["concentration","gini"]}},"required":["op"]},"then":{"anyOf":[{"required":["column"]},{"required":["by"]}]}},{"if":{"properties":{"op":{"const":"pivot"}},"required":["op"]},"then":{"required":["rows_col","cols_col"]}},{"if":{"properties":{"op":{"enum":["trend","rolling","seasonality"]}},"required":["op"]},"then":{"required":["date_col"]}},{"if":{"properties":{"op":{"const":"rolling"}},"required":["op"]},"then":{"required":["window"]}},{"if":{"properties":{"op":{"const":"seasonality"}},"required":["op"]},"then":{"properties":{"grain":{"enum":["month","quarter"]}}}},{"if":{"properties":{"op":{"const":"percentile"}},"required":["op"]},"then":{"required":["column","q"]}},{"if":{"properties":{"op":{"const":"cohort"}},"required":["op"]},"then":{"required":["id_col","date_col"]}},{"if":{"properties":{"op":{"const":"correlation_matrix"}},"required":["op"]},"then":{"required":["columns"]}},{"if":{"properties":{"op":{"const":"join_on"}},"required":["op"]},"then":{"required":["on"]}},{"if":{"properties":{"op":{"const":"compare_series"}},"required":["op"]},"then":{"anyOf":[{"required":["date_col","a_value","b_value"]},{"required":["left","right"]}]}},{"if":{"properties":{"op":{"enum":["numeric_summary","outliers_iqr","currency_mix"]}},"required":["op"]},"then":{"required":["column"]}},{"if":{"properties":{"op":{"const":"filter_rows"}},"required":["op"]},"then":{"required":["filters"]}}],"additionalProperties":false}}
+{"$schema":"https://json-schema.org/draft/2020-12/schema","$id":"https://data-toolkit.local/schemas/analysis-plan.schema.json","title":"Data Analyse operations","type":"array","items":{"type":"object","required":["op"],"properties":{"op":{"enum":["numeric_summary","outliers_iqr","breakdown","period_series","ageing","currency_mix","concentration","pivot","distribution","trend","percentile","cohort","correlation_matrix","rolling","gini","seasonality","join_on","compare_series","filter_rows"]},"name":{"type":"string"},"column":{"type":"string","minLength":1},"by":{"type":"string","minLength":1},"value":{"type":"string"},"top":{"type":"integer","minimum":1},"top_n":{"type":"integer","minimum":1},"k":{"type":["number","string"]},"cap":{"type":"integer","minimum":1},"date_col":{"type":"string","minLength":1},"grain":{"enum":["month","quarter","year"]},"dayfirst":{"type":"boolean"},"as_of":{"type":"string","minLength":1},"buckets":{"type":"array","minItems":1,"items":{"type":"integer","minimum":0}},"rows_col":{"type":"string","minLength":1},"cols_col":{"type":"string","minLength":1},"aggfunc":{"enum":["sum","count","mean"]},"q":{"oneOf":[{"type":"number","minimum":0,"maximum":1},{"type":"array","minItems":1,"items":{"type":"number","minimum":0,"maximum":1}}]},"id_col":{"type":"string","minLength":1},"columns":{"type":"array","minItems":2,"items":{"type":"string","minLength":1}},"window":{"type":"integer","minimum":1},"func":{"enum":["mean","sum","median"]},"on":{"oneOf":[{"type":"string","minLength":1},{"type":"array","minItems":1,"items":{"type":"string","minLength":1}}]},"how":{"enum":["inner","left"]},"a_value":{"type":"string","minLength":1},"b_value":{"type":"string","minLength":1},"a_label":{"type":"string","minLength":1},"b_label":{"type":"string","minLength":1},"left":{"type":"object","required":["date_col"],"properties":{"date_col":{"type":"string","minLength":1},"value":{"type":"string"},"dayfirst":{"type":"boolean"}},"additionalProperties":false},"right":{"type":"object","required":["date_col"],"properties":{"date_col":{"type":"string","minLength":1},"value":{"type":"string"},"dayfirst":{"type":"boolean"}},"additionalProperties":false},"filters":{"type":"array","minItems":1,"items":{"type":"object","required":["op"],"properties":{"col":{"type":"string","minLength":1},"column":{"type":"string","minLength":1},"op":{"type":"string"},"value":{},"values":{"type":"array"},"lo":{},"hi":{}},"additionalProperties":false}}},"allOf":[{"if":{"properties":{"op":{"enum":["numeric_summary","outliers_iqr","currency_mix","distribution"]}},"required":["op"]},"then":{"required":["column"]}},{"if":{"properties":{"op":{"const":"breakdown"}},"required":["op"]},"then":{"required":["by"]}},{"if":{"properties":{"op":{"const":"period_series"}},"required":["op"]},"then":{"required":["date_col"]}},{"if":{"properties":{"op":{"const":"ageing"}},"required":["op"]},"then":{"required":["date_col","as_of"]}},{"if":{"properties":{"op":{"enum":["concentration","gini"]}},"required":["op"]},"then":{"anyOf":[{"required":["column"]},{"required":["by"]}]}},{"if":{"properties":{"op":{"const":"pivot"}},"required":["op"]},"then":{"required":["rows_col","cols_col"]}},{"if":{"properties":{"op":{"enum":["trend","rolling","seasonality"]}},"required":["op"]},"then":{"required":["date_col"]}},{"if":{"properties":{"op":{"const":"rolling"}},"required":["op"]},"then":{"required":["window"]}},{"if":{"properties":{"op":{"const":"seasonality"}},"required":["op"]},"then":{"properties":{"grain":{"enum":["month","quarter"]}}}},{"if":{"properties":{"op":{"const":"percentile"}},"required":["op"]},"then":{"required":["column","q"]}},{"if":{"properties":{"op":{"const":"cohort"}},"required":["op"]},"then":{"required":["id_col","date_col"]}},{"if":{"properties":{"op":{"const":"correlation_matrix"}},"required":["op"]},"then":{"required":["columns"]}},{"if":{"properties":{"op":{"const":"join_on"}},"required":["op"]},"then":{"required":["on"]}},{"if":{"properties":{"op":{"const":"compare_series"}},"required":["op"]},"then":{"anyOf":[{"required":["date_col","a_value","b_value"]},{"required":["left","right"]}]}},{"if":{"properties":{"op":{"const":"filter_rows"}},"required":["op"]},"then":{"required":["filters"]}}],"additionalProperties":false}}

skills/data-visualise/scripts/viz.py

@@ -1864,6 +1864,21 @@ def _render(spec: dict) -> str:
                              show_last=spec.get("show_last", True), unit=spec.get("unit", ""))
         if kind == "waterfall":
             return waterfall(spec.get("steps", []), title=spec.get("title"), unit=spec.get("unit", ""))
+        if kind == "scatter_chart":
+            return scatter_chart(spec.get("x", []), spec.get("y", []),
+                                 title=spec.get("title"),
+                                 x_label=spec.get("x_label"), y_label=spec.get("y_label"),
+                                 unit_x=spec.get("unit_x", ""),
+                                 unit_y=spec.get("unit_y", spec.get("unit", "")),
+                                 labels=spec.get("labels"),
+                                 trend_line=spec.get("trend_line", False))
+        if kind == "histogram":
+            return histogram(spec.get("values", []), bins=spec.get("bins", 10),
+                             title=spec.get("title"), unit=spec.get("unit", ""))
+        if kind == "stacked_bar":
+            return stacked_bar(spec.get("data", {}), title=spec.get("title"),
+                               unit=spec.get("unit", ""),
+                               legend=spec.get("legend", True))
         if kind == "section":
             return section(spec.get("title", ""), *[_render(c) for c in spec.get("blocks", [])])
         if kind == "grid":

tests/test_engine.py

@@ -746,6 +746,10 @@ def test_suggest_blocks_proposes_stacked_bar_for_pivot():
     block_types = [b["type"] for b in specs[0]["blocks"]]
     assert "heatmap" in block_types
     assert "stacked_bar" in block_types, f"expected stacked_bar, got {block_types}"
+    # suggest proposes it; blocks_from_analysis must actually render it (merge
+    # follow-up: the convenience helper lagged _viz_block / suggest).
+    html = viz.blocks_from_analysis(analysis)
+    assert any("<rect" in h for h in html)
 
 
 def test_suggest_blocks_proposes_histogram_for_distribution_when_values_carried():
@@ -758,6 +762,8 @@ def test_suggest_blocks_proposes_histogram_for_distribution_when_values_carried(
     specs = viz.suggest_blocks_from_analysis(analysis)
     block_types = [b["type"] for b in specs[0]["blocks"]]
     assert "histogram" in block_types, f"expected histogram, got {block_types}"
+    html = viz.blocks_from_analysis(analysis)
+    assert any("value" in h and "<rect" in h for h in html)
     # without values, no histogram is proposed
     analysis2 = {"results": [{"op": "distribution", "name": "x", "result": {
         "skewness": 1.2, "kurtosis": 2.1, "classification": "right-skewed"}}]}