BetaLenzon is in beta — the Free tier is 100% free while we're in beta (bring your own Anthropic key, public repos).See plans →
Lenzon

msitarzewski/agency-agents PR #729 — Add Data Visualization Engineer specialist — PR #729

msitarzewski/agency-agents · pull request #729 ·

Loading…

Transcript

PlainEnglish

We're looking at pull request seven twenty-nine. This PR brings a new specialist agent to the agency-agents repository — a Data Visualization Engineer. A hundred fifty-two new lines across two files, no deletions. It's pure addition, extending the catalog with a visualization-focused expert.

PlainEnglish

Two main goals here. First, create the Data Visualization Engineer specialist agent with all the guidance needed for chart selection, accessibility, and performance. Second, register the agent in the README so users can discover it. And we're keeping the existing catalog structure intact — no disruptions to the patterns already in place.

PlainEnglish

The motivation is straightforward — the repository is a curated collection of AI specialist agents covering technical domains. Before this PR, there was no visualization expert. This change fills that gap, bringing guidance on a discipline that bridges design, perception science, and performance engineering.

Architecture

Let's see where the changes live. One new file under the engineering directory — the data visualization engineer agent spec, a hundred fifty-one lines of structured guidance. And one updated file — the README, which gets a single new table row pointing to the agent. Clean and focused.

Architecture

Here's the heart of the agent spec. YAML frontmatter at the top defines identity — name, description, color, emoji, and a vibe statement. Then critical rules: chart-type selection before aesthetics, perceptual honesty with zero baselines, colorblind-safe palettes tested for the eight percent of viewers with color vision deficiency, and performance budgets that guide when to use SVG, Canvas, or WebGL.

Architecture

The agent delivers production-ready guidance across five key areas. A chart-type selection table that maps data structure and question type to visual encodings. A perceptual honesty checklist ranking position, length, angle, area, and hue by quantitative precision. Colorblind-safe code examples for D3, Vega, and Canvas, grounded in the eight-percent CVD prevalence. Performance thresholds that tell you when to reach for SVG, Canvas, or WebGL. And an eight-step workflow from data audit through accessibility validation to deployment.

Architecture

Now the README update. A single table row slips in between the WordPress Shopping Cart Engineer and the Payments and Billing Engineer. Emoji, linked title, key topics — chart-type selection, perceptually honest encodings, colorblind-safe palettes, performance — and a three-word summary. The table structure and link format stay consistent with the existing catalog.

PlainEnglish

Here's the outcome. Before this PR, the repository had thirteen engineering specialists and a domain gap where visualization guidance should be. After, we have fourteen agents. The Data Visualization Engineer is discoverable in the README index, ready to guide users through chart selection, colorblind-safe design, and performance optimization across D3, Canvas, and WebGL.

PlainEnglish

We looked at YAML frontmatter structure, markdown formatting, README table insertion position, link validity, technical content accuracy — chart mappings and CVD prevalence grounded in established science — header hierarchy, and file endings. Everything checked out. No concerns to flag.

PlainEnglish

That's pull request seven twenty-nine. A clean catalog extension that adds visualization expertise without disrupting existing patterns. The agent spec lives at engineering slash engineering-data-visualization-engineer dot md, the README table entry at line one fifty-two, and a hundred fifty-one lines of guidance on chart selection, accessibility, and performance are now part of the catalog.

How this was made

Lenzon read msitarzewski/agency-agents at pull request #729 and generated this walkthrough automatically. The narration above is the transcript of what it says.

Explain a pull request from your own repo

Point Lenzon at a repo or a pull request and get a narrated walkthrough like this one.

Try it