narrative-text-visualization_skill

Overview

This skill generates structured narrative text visualizations from data using T8 Syntax. It turns raw metrics and qualitative observations into annotated, LLM-friendly narratives that render as readable reports with embedded mini-charts and semantic entities. Use it to produce interpretable data stories for web apps or documentation.

How this skill works

You provide the analysis requirements and source data, and the skill outputs T8 Syntax—a Markdown-like, declarative language with entity annotations (e.g., metric_value, ratio_value, dim_value). The output can include metadata (origin, unit, assessment, detail) so frontends can render visual cues and mini visualizations. The skill also generates example HTML/React/Vue snippets to render the T8 content.

When to use it

• Create data interpretation reports, executive summaries, or structured articles from verified data.
• Annotate metrics and trends with semantic metadata for downstream visualization or automated styling.
• Generate LLM-friendly prompts or templates that produce consistent narrative outputs.
• Embed concise, styled narrative blocks in web dashboards using HTML, React, or Vue.
• Produce reproducible narrative templates that require programmatic rendering of entity-driven mini-charts.

Best practices

• Always source numbers from authoritative, public sources (financial reports, Reuters, IDC).
• Annotate every meaningful data point with an appropriate entity type and origin value.
• Use assessment metadata (positive/negative/equal/neutral) to enable consistent visual indicators.
• Include detail arrays for rank, difference, anomaly, association, distribution, and seasonality types.
• Keep narrative structure clear: headings, paragraphs, and lists for readable rendering.

Example use cases

• Quarterly business reviews that combine KPI callouts with short explanatory paragraphs and trend chips.
• Market reports where regional breakdowns use dim_value, contribute_ratio, and rank entities for leaderboards.
• Product analytics summaries that annotate conversion rates, churn deltas, and anomaly bursts for engineers and PMs.
• Automated content generation pipelines where an LLM outputs T8 syntax for downstream rendering in a dashboard.
• Documentation pages that need compact, styled data narratives with embedded small charts and semantic tags.