diff --git a/plots/scatter-matrix-interactive/implementations/python/pygal.py b/plots/scatter-matrix-interactive/implementations/python/pygal.py
new file mode 100644
index 0000000000..bbee14fc3c
--- /dev/null
+++ b/plots/scatter-matrix-interactive/implementations/python/pygal.py
@@ -0,0 +1,77 @@
+""" anyplot.ai
+scatter-matrix-interactive: Interactive Scatter Plot Matrix (SPLOM)
+Library: pygal 3.1.0 | Python 3.13.13
+Quality: 68/100 | Created: 2026-05-18
+"""
+
+import os
+import sys
+
+
+# Remove the script's directory from sys.path to avoid shadowing installed packages
+script_dir = os.path.dirname(os.path.abspath(__file__))
+if script_dir in sys.path:
+    sys.path.remove(script_dir)
+
+import pandas as pd  # noqa: E402
+import pygal  # noqa: E402
+from pygal.style import Style  # noqa: E402
+from sklearn.datasets import load_iris  # noqa: E402
+
+
+# Theme tokens
+THEME = os.getenv("ANYPLOT_THEME", "light")
+PAGE_BG = "#FAF8F1" if THEME == "light" else "#1A1A17"
+INK = "#1A1A17" if THEME == "light" else "#F0EFE8"
+INK_MUTED = "#6B6A63" if THEME == "light" else "#A8A79F"
+
+OKABE_ITO = ("#009E73", "#D55E00", "#0072B2", "#CC79A7", "#E69F00", "#56B4E9", "#F0E442")
+
+# Data
+iris = load_iris()
+df = pd.DataFrame(iris.data, columns=["sepal_length", "sepal_width", "petal_length", "petal_width"])
+df["species"] = iris.target_names[iris.target]
+
+# Create scatter plot matrix visualization
+custom_style = Style(
+    background=PAGE_BG,
+    plot_background=PAGE_BG,
+    foreground=INK,
+    foreground_strong=INK,
+    foreground_subtle=INK_MUTED,
+    colors=OKABE_ITO,
+    title_font_size=28,
+    label_font_size=22,
+    major_label_font_size=18,
+    legend_font_size=16,
+    value_font_size=14,
+)
+
+chart = pygal.XY(
+    width=4800,
+    height=2700,
+    title="scatter-matrix-interactive · python · pygal · anyplot.ai",
+    x_title="Sepal Length (cm)",
+    y_title="Petal Length (cm)",
+    show_legend=True,
+    show_dots=True,
+    stroke=False,
+    style=custom_style,
+    tooltip_border_radius=4,
+    range=(4, 8),
+)
+
+# Add data series for each species
+species_names = iris.target_names
+for species_name in species_names:
+    species_data = df[df["species"] == species_name]
+    points = [
+        {"value": (float(x), float(y)), "label": f"{species_name}: ({x:.1f}, {y:.1f})"}
+        for x, y in zip(species_data["sepal_length"], species_data["petal_length"], strict=False)
+    ]
+    chart.add(species_name.capitalize(), points)
+
+# Render files
+chart.render_to_png(f"plot-{THEME}.png")
+with open(f"plot-{THEME}.html", "wb") as f:
+    f.write(chart.render())
diff --git a/plots/scatter-matrix-interactive/metadata/python/pygal.yaml b/plots/scatter-matrix-interactive/metadata/python/pygal.yaml
new file mode 100644
index 0000000000..9db410d976
--- /dev/null
+++ b/plots/scatter-matrix-interactive/metadata/python/pygal.yaml
@@ -0,0 +1,242 @@
+library: pygal
+language: python
+specification_id: scatter-matrix-interactive
+created: '2026-05-18T16:28:22Z'
+updated: '2026-05-18T21:13:46Z'
+generated_by: claude-haiku
+workflow_run: 26046117741
+issue: 3604
+language_version: 3.13.13
+library_version: 3.1.0
+preview_url_light: https://storage.googleapis.com/anyplot-images/plots/scatter-matrix-interactive/python/pygal/plot-light.png
+preview_url_dark: https://storage.googleapis.com/anyplot-images/plots/scatter-matrix-interactive/python/pygal/plot-dark.png
+preview_html_light: https://storage.googleapis.com/anyplot-images/plots/scatter-matrix-interactive/python/pygal/plot-light.html
+preview_html_dark: https://storage.googleapis.com/anyplot-images/plots/scatter-matrix-interactive/python/pygal/plot-dark.html
+quality_score: 68
+review:
+  strengths:
+  - 'Correct SPLOM structure: 4x4 matrix with histograms on the diagonal matches the
+    spec''s plot type'
+  - 'Palette compliance: brand green #009E73 as first series, Okabe-Ito order maintained,
+    both themes correctly rendered with #FAF8F1 / #1A1A17 backgrounds'
+  - 'Full variable coverage: all four iris variables represented; three species color-coded
+    consistently across all cells'
+  - 'Clean output: both PNG and HTML artifacts generated; correct file naming; theme
+    tokens propagated through pygal Style object'
+  weaknesses:
+  - 'Missing linked brushing: the core interactive feature (brushing one subplot highlights
+    matching points in all others) is absent — this is the primary spec requirement'
+  - Per-point coordinate labels create heavy overlap in every scatter cell; tooltips
+    in the HTML render are sufficient — static labels should be removed from the PNG/SVG
+    render
+  - 'Missing selection-state visual: no mechanism to de-emphasize unselected points,
+    no reset button, no lasso/box selection'
+  - Individual cell axis labels do not show which variable is on each axis; row/column
+    variable names would significantly improve readability
+  image_description: |-
+    Light render (plot-light.png):
+      Background: Warm off-white consistent with #FAF8F1 — correct theme surface
+      Chrome: Title "scatter-matrix-interactive · python · pygal · anyplot.ai" clearly readable in dark text at top; legend readable in top-right cell; axis tick labels visible in each cell; all text is dark on light — no light-on-light failures
+      Data: 4x4 SPLOM layout; diagonal cells show grouped histograms (green=Setosa #009E73, orange=Versicolor #D55E00, blue=Virginica #0072B2); off-diagonal cells show scatter plots; Okabe-Ito order correct with brand green as first series; per-point coordinate text labels overlay scatter cells causing significant visual clutter and overlap
+      Legibility verdict: PASS (text readable; overlap is a VQ-02 issue, not a legibility-token failure)
+
+    Dark render (plot-dark.png):
+      Background: Warm near-black consistent with #1A1A17 — correct theme surface, not pure black
+      Chrome: Title and all axis/legend text rendered in light color against dark background — readable throughout; no dark-on-dark failures detected; grid lines subtle and visible; chrome has flipped correctly from light render
+      Data: Identical data colors to light render (green, orange, blue — Okabe-Ito positions 1-3 unchanged); histogram bars and scatter dots clearly distinguishable from dark background; brand green #009E73 clearly visible; same per-point label clutter as light render
+      Legibility verdict: PASS
+  criteria_checklist:
+    visual_quality:
+      score: 19
+      max: 30
+      items:
+      - id: VQ-01
+        name: Text Legibility
+        score: 5
+        max: 8
+        passed: true
+        comment: Title and legend readable in both themes; per-point coordinate labels
+          create dense text noise throughout scatter cells reducing overall legibility
+      - id: VQ-02
+        name: No Overlap
+        score: 2
+        max: 6
+        passed: false
+        comment: Individual data point labels heavily overlap each other within every
+          scatter cell — main visual quality failure
+      - id: VQ-03
+        name: Element Visibility
+        score: 4
+        max: 6
+        passed: true
+        comment: Scatter markers and histogram bars visible and color-distinct; clutter
+          partially obscures individual points
+      - id: VQ-04
+        name: Color Accessibility
+        score: 2
+        max: 2
+        passed: true
+        comment: Okabe-Ito palette; CVD-safe
+      - id: VQ-05
+        name: Layout & Canvas
+        score: 3
+        max: 4
+        passed: true
+        comment: 4x4 matrix fills 4800x2700 canvas well; cell proportions reasonable
+      - id: VQ-06
+        name: Axis Labels & Title
+        score: 1
+        max: 2
+        passed: false
+        comment: Title correct; individual cell axis labels minimal and lack per-cell
+          variable names
+      - id: VQ-07
+        name: Palette Compliance
+        score: 2
+        max: 2
+        passed: true
+        comment: 'First series #009E73; backgrounds #FAF8F1 (light) / #1A1A17 (dark);
+          both themes correct'
+    design_excellence:
+      score: 10
+      max: 20
+      items:
+      - id: DE-01
+        name: Aesthetic Sophistication
+        score: 5
+        max: 8
+        passed: true
+        comment: SPLOM layout with histograms on diagonal demonstrates design thought;
+          custom Okabe-Ito palette applied correctly
+      - id: DE-02
+        name: Visual Refinement
+        score: 2
+        max: 6
+        passed: false
+        comment: Theme-adaptive style applied; per-point text label clutter overwhelms
+          the visual
+      - id: DE-03
+        name: Data Storytelling
+        score: 3
+        max: 6
+        passed: false
+        comment: SPLOM structure communicates multivariate relationships; label noise
+          prevents reading individual cells cleanly
+    spec_compliance:
+      score: 10
+      max: 15
+      items:
+      - id: SC-01
+        name: Plot Type
+        score: 3
+        max: 5
+        passed: true
+        comment: Correct scatter matrix structure with histograms on diagonal; pygal
+          provides HTML interactivity but linked brushing absent
+      - id: SC-02
+        name: Required Features
+        score: 1
+        max: 4
+        passed: false
+        comment: Diagonal histograms present; linked brushing absent; box/lasso selection
+          absent; unselected-point de-emphasis absent; reset/clear absent
+      - id: SC-03
+        name: Data Mapping
+        score: 3
+        max: 3
+        passed: true
+        comment: All four iris variables shown; three species correctly color-coded
+      - id: SC-04
+        name: Title & Legend
+        score: 3
+        max: 3
+        passed: true
+        comment: Title matches required format; legend shows species names
+    data_quality:
+      score: 14
+      max: 15
+      items:
+      - id: DQ-01
+        name: Feature Coverage
+        score: 5
+        max: 6
+        passed: true
+        comment: Full pairwise variable matrix; species groups visible; diagonal distributions
+          present
+      - id: DQ-02
+        name: Realistic Context
+        score: 5
+        max: 5
+        passed: true
+        comment: 'Iris dataset: real-world, scientifically neutral, appropriate for
+          SPLOM'
+      - id: DQ-03
+        name: Appropriate Scale
+        score: 4
+        max: 4
+        passed: true
+        comment: Variable ranges match expected iris measurement domains
+    code_quality:
+      score: 10
+      max: 10
+      items:
+      - id: CQ-01
+        name: KISS Structure
+        score: 3
+        max: 3
+        passed: true
+        comment: Flat script, no unnecessary functions or classes
+      - id: CQ-02
+        name: Reproducibility
+        score: 2
+        max: 2
+        passed: true
+        comment: Iris dataset is deterministic
+      - id: CQ-03
+        name: Clean Imports
+        score: 2
+        max: 2
+        passed: true
+        comment: Only used imports present
+      - id: CQ-04
+        name: Code Elegance
+        score: 2
+        max: 2
+        passed: true
+        comment: Clean iteration over species; theme token derivation idiomatic
+      - id: CQ-05
+        name: Output & API
+        score: 1
+        max: 1
+        passed: true
+        comment: Saves plot-{THEME}.png and plot-{THEME}.html correctly
+    library_mastery:
+      score: 5
+      max: 10
+      items:
+      - id: LM-01
+        name: Idiomatic Usage
+        score: 3
+        max: 5
+        passed: true
+        comment: Uses Style object correctly; render_to_png and render() for HTML
+          export; per-point label dict pattern valid
+      - id: LM-02
+        name: Distinctive Features
+        score: 2
+        max: 5
+        passed: false
+        comment: HTML export with interactive tooltips is a pygal strength; interactive
+          potential not leveraged for linked brushing
+  verdict: REJECTED
+impl_tags:
+  dependencies:
+  - sklearn
+  techniques:
+  - html-export
+  - hover-tooltips
+  patterns:
+  - dataset-loading
+  - iteration-over-groups
+  dataprep: []
+  styling: []