talont-org/vis-lens-color-access

Chromatic Accessibility Visualization Lens

Philosophical Mode: Chromatic Primary Question: "Is the color encoding accessible and perceptually uniform?" Focus: Colorblind Safety (Okabe-Ito, Paul Tol palettes), Perceptual Uniformity (viridis/cividis pass; jet/rainbow fail), Non-Color Redundant Encoding (shape + line-style), Text Size Minimums

Arguments

/autoskillit:vis-lens-color-access [context_path] [experiment_plan_path]

• context_path (optional positional arg 1) — Absolute path to a lens context file containing IV/DV tables, H0/H1 hypotheses, controlled variables, and success criteria. If provided, read this file before beginning analysis to obtain structured context. If omitted, discover context by exploring the CWD.
• experiment_plan_path (optional positional arg 2) — Absolute path to the full experiment plan. If provided, read for complete experimental methodology and design. If omitted, locate the experiment plan by exploring the CWD.

When to Use

• Auditing figures for colorblind accessibility before submission
• Checking whether colormaps are perceptually uniform for quantitative data
• Verifying that color-distinguished series have redundant non-color encodings
• Reviewing font sizes for accessibility and publication standards
• User invokes /autoskillit:vis-lens-color-access

Critical Constraints

NEVER:

• Modify any source code files
• Do not litter the codebase with useless comments, TODO markers, or explanatory annotations — the skill output and diagram speak for themselves
• Create files outside {{AUTOSKILLIT_TEMP}}/vis-lens-color-access/
• Use jet or rainbow colormaps for any quantitative data — these are perceptually non-uniform and fail colorblind simulation
• Encode quantitative data with hue alone (no luminance gradient)

ALWAYS:

• Use Okabe-Ito (8 colors) or Paul Tol palettes for categorical nominal data

• Use viridis, cividis, or mako for sequential/diverging quantitative data

• Add a redundant non-color encoding (shape marker AND/OR line-style dash pattern) for every color-distinguished series in line and scatter plots

• Caption text and axis labels: minimum 8pt in final figure; 10pt preferred for publication

• BEFORE creating any diagram, LOAD the /autoskillit:mermaid skill using the Skill tool — this is MANDATORY

• If the Skill tool cannot be used (disable-model-invocation) or refuses this invocation, do NOT proceed with diagram creation. Abort this step and omit the diagram from output.

• Write output to {{AUTOSKILLIT_TEMP}}/vis-lens-color-access/vis_spec_color_access_{YYYY-MM-DD_HHMMSS}.md (relative to the current working directory)

• After writing the file, emit the structured output token as literal plain text with no markdown formatting on the token name (the adjudicator performs a regex match):

  diagram_path = /absolute/path/to/{{AUTOSKILLIT_TEMP}}/vis-lens-color-access/vis_spec_color_access_{...}.md
  

---

Analysis Workflow

Step 0: Parse optional arguments

If positional arg 1 (context_path) is provided and the file exists, read it to obtain IV/DV tables, H0/H1 hypotheses, controlled variables, and success criteria. If positional arg 2 (experiment_plan_path) is provided and exists, read the experiment plan for full methodology. Use this structured context as the foundation for Steps 1–4; skip the CWD exploration for these fields if the context file supplies them.

Step 1: Inventory Color Usage

Scan experiment plan, context file, and codebase for:

Palette and Colormap Names

• Find all palette names and colormap strings in plotting code
• Look for: cmap=, palette=, color=, colors=, sns.color_palette, plt.cm.
• Flag FAIL: jet, rainbow, hsv, Spectral, hot, cool
• Pass: viridis, cividis, mako, okabe-ito, wong, tol, tab10

Hue-Only Encoding

• Detect cases where quantitative data uses hue without luminance gradient
• Look for: diverging colormaps applied to continuous scales, categorical palettes applied to ordinal data

Series Count

• Count the number of color-distinguished series per figure
• Look for: legend entries, label= parameters, series lists

Step 2: Redundant Encoding Audit

For each color-distinguished series in a line or scatter plot:

• Check whether a marker shape (marker=) is also assigned
• Check whether a line-style dash (linestyle=, dashes=) is also assigned
• If neither is present: flag as WARNING — "Series relies on color alone; add marker or linestyle"

Step 3: Text Size Audit

Scan figure creation calls for font size parameters:

• Look for: fontsize=, labelsize=, title_fontsize=, tick_params(labelsize=...), rcParams
• Flag any value < 8 as WARNING: "Font size {N}pt is below 8pt minimum"
• Flag any value < 10 as INFO: "Font size {N}pt is below 10pt preferred for publication"

Step 4: Emit yaml:figure-spec Blocks

For each figure, emit one yaml:figure-spec fenced block with palette field filled. Then LOAD /autoskillit:mermaid and create a diagram showing palette audit → redundant encoding → text audit → verdict.

---

Output Template

# Chromatic Accessibility Spec: {System / Experiment Name}

**Lens:** Chromatic Accessibility (Chromatic)
**Question:** Is the color encoding accessible and perceptually uniform?
**Date:** {YYYY-MM-DD}
**Scope:** {What was analyzed}

## Color Audit Summary

| Figure | Palette | Colorblind Safe | Redundant Encoding | Min Font | Status |
|--------|---------|-----------------|-------------------|----------|--------|
| {fig-01} | jet | FAIL | no | 7pt | FAIL |
| {fig-02} | okabe-ito | PASS | yes (marker+dash) | 10pt | OK |

## Figure Specs

```yaml
# yaml:figure-spec — canonical schema (spec_version: "1.0")
figure_id: "fig-02-ablation-comparison"
figure_title: "Ablation: Component Contribution"
spec_version: "1.0"
chart_type: "line"
chart_type_fallback: "grouped-bar"
perceptual_justification: "Okabe-Ito is colorblind-safe; marker shapes provide redundant encoding."
data_source: "results/ablation.csv"
data_mapping:
  x: "epoch"
  y: "accuracy"
  color: "variant"
  size: ""
  facet: ""
layout:
  width_inches: 5.0
  height_inches: 3.5
  dpi: 300
stat_overlay:
  type: "error_bar"
  measure: "CI95"
  n_seeds: 5
annotations: ["okabe-ito palette; marker shapes assigned; 10pt axis labels"]
anti_patterns: ["ap-hue-only-encoding"]
palette: "okabe-ito"
format: "pdf"
target_dpi: 300
library: "matplotlib"
report_section: "Section 5 Ablation"
priority: "P1"
placement_tier: "main"
conflicts: []
metadata:
  created_by: "vis-lens-color-access"
  reviewed_by: ""
  last_updated: "{YYYY-MM-DD}"

Chromatic Accessibility Diagram

%%{init: {'flowchart': {'nodeSpacing': 50, 'rankSpacing': 60, 'curve': 'basis'}}}%%
flowchart TB
    %% CLASS DEFINITIONS %%
    classDef cli fill:#1a237e,stroke:#7986cb,stroke-width:2px,color:#fff;
    classDef stateNode fill:#004d40,stroke:#4db6ac,stroke-width:2px,color:#fff;
    classDef handler fill:#e65100,stroke:#ffb74d,stroke-width:2px,color:#fff;
    classDef output fill:#00695c,stroke:#4db6ac,stroke-width:2px,color:#fff;
    classDef detector fill:#b71c1c,stroke:#ef5350,stroke-width:2px,color:#fff;

    subgraph Palette ["PALETTE AUDIT"]
        P1["palette = {name}<br/>━━━━━━━━━━<br/>colorblind safe: PASS / FAIL"]
    end

    subgraph Redundant ["REDUNDANT ENCODING"]
        R1["marker: yes/no<br/>linestyle: yes/no<br/>━━━━━━━━━━<br/>series count = {N}"]
    end

    subgraph Text ["TEXT SIZE AUDIT"]
        T1["min font = {N}pt<br/>━━━━━━━━━━<br/>OK / WARNING / INFO"]
    end

    subgraph Verdict ["VERDICT"]
        V1["{OK / WARNING / FAIL}<br/>━━━━━━━━━━<br/>{reason}"]
    end

    P1 --> R1
    R1 --> T1
    T1 --> V1

    class P1 stateNode;
    class R1 handler;
    class T1 cli;
    class V1 detector;

Color Legend: | Color | Category | Description | |-------|----------|-------------| | Dark Teal | Palette | Colorblind safety audit result | | Orange | Redundant Encoding | Marker and linestyle check | | Dark Blue | Text Size | Font size audit result | | Red | Verdict | OK / WARNING / FAIL assessment |

---

## Pre-Diagram Checklist

Before creating the diagram, verify:

- [ ] LOADED `/autoskillit:mermaid` skill using the Skill tool
- [ ] Using ONLY classDef styles from the mermaid skill (no invented colors)
- [ ] Diagram will include a color legend table
- [ ] Every failing colormap (jet, rainbow) is explicitly flagged as FAIL
- [ ] Every color-distinguished series has been checked for redundant encoding
- [ ] Every font size below 8pt has been flagged as WARNING