| name | dataviz-planner |
| description | Plan charts, figures, and visualizations — data mapping, chart selection, interaction design. |
| version | 0.1.0 |
Data Visualization Planner
Planner skill for designing data visualizations before you create them.
Use this skill to plan visualizations that answer specific questions, fit your audience's literacy level, and make disparities visible rather than hiding them in aggregates.
JTBD (Jobs To Be Done)
Primary Job
When I need to create a chart or visualization but haven't yet resolved what question it answers, what data pattern it shows, or whether my audience can interpret it correctly,
I want a concrete visualization design plan before I start building,
so I can avoid creating something that looks polished but misleads viewers or hides the patterns that actually matter.
Secondary Jobs
- When the audience for a visualization has a different data literacy level than I assumed — executives who can't read confidence intervals, or a general public audience that will misread a logarithmic scale — I want the encoding choices matched to who will actually see it, so the chart informs rather than confuses.
- When a dataviz-critic review returned REVISE or REJECT and I need to redesign rather than just patch, I want a redesign plan that addresses the root encoding or framing failures, so I don't rebuild the same mistake with better typography.
Job Layers
- Functional: Resolve chart type selection, data encoding strategy, audience literacy match, accessibility requirements, and equity implications before any tool is opened or code is written.
- Emotional: Reduce the anxiety of committing to a chart type or design approach and discovering after publication that it was the wrong one for the data pattern or audience.
- Social: Helps the user present a defensible design rationale to editors, reviewers, or stakeholders — "this is why we chose a small-multiples layout over a single stacked bar" — backed by planning rather than instinct.
This Skill Is For
- A user about to create a chart or visualization who hasn't yet answered: what question does this answer, for whom, and which encoding suits the data pattern?
- A user working with disaggregated or equity-sensitive data who needs to plan how to make disparities visible rather than letting them vanish into an aggregate.
- A user building multiple visualizations for a report, dashboard, or presentation who needs consistent encoding choices and a coherent visual language before creating any individual chart.
- A user whose previous visualization received a REVISE or REJECT verdict and needs a redesign plan that goes back to the question and encoding level, not just surface fixes.
This Skill Is NOT For
- A user with an existing chart who primarily needs a quality verdict on what was built; use
dataviz-critic instead.
- A user whose unresolved problem is end-to-end dashboard architecture and KPI hierarchy rather than individual chart design; use
dashboard-planner instead.
Paired With
dataviz-critic: After the chart or visualization is created, use it to audit whether the encoding choices are honest, accessible, and effective.
dashboard-planner: Use this when the unresolved problem is the architecture of a multi-chart dashboard rather than the design of any individual visualization within it.
Resolution Paths
| User's Situation | What Happens | What They Leave With |
|---|
| Has data and a goal but no chart type or encoding plan | The skill resolves the question, data pattern, chart type selection, and encoding strategy before creation begins | A concrete design specification: chart type, axes, encoding choices, accessibility requirements |
| Has an audience literacy or context mismatch (complex data, general audience) | The skill identifies the mismatch and adapts the encoding, annotation, and framing choices to the actual audience | A design approach calibrated to what this audience can reliably interpret |
| Has a failed visualization after a REVISE or REJECT verdict | The skill diagnoses the root encoding or framing failure and produces a redesign plan | A redesign plan addressing root causes, not surface cosmetics |
When to Escalate
- If the user already has a built visualization and needs a quality verdict rather than a design plan, escalate to
dataviz-critic.
- If the user's problem is the end-to-end architecture of a dashboard system rather than the design of individual charts, escalate to
dashboard-planner.
Purpose
Plan data visualizations strategically, not reactively:
- Clarify the question: What question does this visualization answer?
- Know your audience: Who is viewing it? What's their data literacy?
- Assess the data: Is your data complete? Are sample sizes adequate? Can you disaggregate by population groups?
- Choose the right chart type: Comparison vs. trend vs. distribution vs. composition — each needs different visualization
- Design for accessibility: Color-blind safe palettes, patterns for colorblind users, clear annotations
- Make equity visible: Disaggregate data by population; prevent disparities from hiding in aggregates
This skill produces a detailed plan that guides actual visualization creation.
Use_When
- Planning a visualization before starting design (planning vs. reactively creating charts)
- Building dashboards, reports, or presentations with multiple visualizations
- You need to disaggregate data by population and show disparities
- Determining what visualization type answers your question best
- Ensuring visualizations are accessible to all users (colorblind, low vision, etc.)
- Creating publication-quality visualizations that will be reviewed
- You want to test whether a visualization effectively communicates its intended message
Do_Not_Use_When
- You've already created a visualization and want critique (use dataviz-critic instead)
- You need quick exploratory charts (this is for intentional design)
- The visualization is a throwaway internal sanity check
- You're just learning data visualization basics (use visualization-standards reference instead)
Companion_Skills
- dataviz-critic: Use AFTER creating visualizations to review design decisions and identify gaps
- stakeholder-report-writer: Use when planning visualizations for reports or presentations
- visualization-standards: Reference for color palettes, chart type conventions, accessibility guidance
Steps
-
Identify the question and data: Provide the research/business question and the data you'll be visualizing
-
Determine audience and context: Clarify:
- Who is the audience? (Clinician, board member, community group, researcher, general public)
- Where will it be displayed? (Report, presentation, dashboard, website, social media, print)
- What decision should it support?
- Data literacy level? (Low, moderate, high)
-
Invoke the dataviz-planner subagent: Delegate to subagent with the full planning protocol below:
- Local routing authority (default): Route through the repository agent named by this skill; if it is unavailable, use a host general-purpose worker with the full protocol. The catalog/meta-router owns route and model selection.
- Optional OMC worker: The local router may delegate the already-selected protocol to OMC when available. OMC does not choose the route or model policy.
-
Return planning output: Present the structured visualization plan with:
- Question and audience clarity
- Data assessment and quality assurance
- Chart type selection with rationale
- Design specifications (color, typography, annotations)
- Equity review and testing plan
The plan guides actual visualization creation. Use dataviz-critic to review the completed visualization.
Tool_Usage
When invoking dataviz-planner:
- Use Read to load data files and understand structure
- Use Grep to verify data fields, completeness, value ranges
- Use Bash to run basic data exploration (row counts, unique values, ranges)
- Understand the data thoroughly before planning chart design
Related_Skills
- dataviz-critic: Post-creation review of visualization design decisions
- stakeholder-report-writer: When planning visualizations for reports
- visualization-standards: Reference for accessibility, color, chart type conventions