Claude got a new Skill - /dataviz

Claude got a new Skill - /dataviz

Introduction to slash datavis: A New Skill in Clod Co

Overview of slash datavis

  • The new skill, slash datavis, is designed to fact-check chart colors using a script rather than relying on visual inspection.
  • It emphasizes the importance of selecting chart forms before assigning colors based on their functional roles: identity, magnitude, or polarity.

Validation Process

  • After color assignment, users must run a validator that assesses the color palette for contrast and color blindness compatibility.
  • The validator provides a pass or fail result along with specific reasons for its assessment.

Real-world Application and Results

Example of Color Validation

  • An example is provided where three colors (blue, violet, orange) were tested in light mode and passed but failed in dark mode due to issues for red-green color blind viewers.
  • The initial failure was attributed to insufficient color distance; swapping violet for aqua resolved the issue, leading to successful validation in both modes.

Impact on Releases

  • The validated colors are used consistently across all Clod Co releases over the past three months, ensuring accessibility for all users regardless of visual impairments.
  • This approach contrasts with traditional methods where personal taste may lead to charts that appear fine but are not universally accessible.

Conclusion and Call to Action

Engaging the Audience

  • A question is posed regarding whether viewers check their charts for color blindness or rely solely on visual judgment.
  • Viewers are encouraged to engage with the content by commenting and exploring more about AI through the dynamist.ai community.
Video description

Claude Code's new /dataviz skill fact-checks your chart colors with a runnable validator — contrast, color-blindness (CVD), and light + dark mode — instead of your eyes. Shipped in Claude Code v2.1.198, it caught a blue-vs-violet collision no designer would have spotted. ---- 🚀 DYNAMOUS AI COMMUNITY Want to learn agentic coding with live daily events and workshops? Check out Dynamous AI: https://dynamous.ai/?code=646a60 Get 10% off here 👉 https://shorturl.smartcode.diy/dynamous_ai_10_percent_discount ⚡ HOSTINGER — RELIABLE HOSTING FOR YOUR PROJECTS (10% OFF) Whether you're shipping a portfolio, a side project, n8n flows, or AI agents — I use Hostinger for fast, affordable VPS + web hosting. Get 10% off here 👉 https://hostinger.com/DIYSMARTCODE (Affiliate link — costs you nothing, supports the channel.) ---- What you will see in this 100-second breakdown of the Claude Code /dataviz skill: - Why /dataviz picks the chart form first and assigns color by its job (identity, magnitude, polarity) — before any color is chosen - The validate_palette.js validator scoring a real palette for contrast + red-green color-blindness, returning pass / fail with the exact reason - A live FAIL: blue and violet collapse to a color-distance (ΔE) of 2.5 in dark mode, where 12 is the floor — swap violet for aqua, and it passes both modes - The payoff: three validated colors rendering every Claude Code release over the last 3 months — features, bugfixes, and improvements, stacked by week, readable in light and dark Claude Code /dataviz skill: run /dataviz inside Claude Code (v2.1.198+) Do you check your charts for color-blindness with a validator, or do you just eyeball the colors and ship? Drop it in the comments. #ClaudeCode #dataviz #DataVisualization #AICoding #Anthropic #ColorBlindness #Accessibility #DevTools #AgenticCoding #ChartDesign #DataViz #A11y #CVD #ClaudeCodeSkills #CodingTools #AItools #DarkMode #FrontendDev #DataScience #SoftwareEngineering