dossier

Provide data from a SharePoint list, uploaded documents, or a verbal description, and the skill builds a validated JSON document then renders it into a self-contained HTML page. Two to eight subjects render as a responsive card grid; a single subject renders as a full one-pager briefing.

You can use SharePoint lists read via list_items, documents or tables pasted in chat, or a purely verbal description. If the data is incomplete, the skill interviews you in a single message about subject type, count, key facts, and palette before building.

Yes, the output is designed for SharePoint sandboxed iframes. It contains no external URLs, CDN links, web fonts, or script src tags, and all values are pre-rendered so the page displays fully even with JavaScript disabled.

Grid mode supports 2 to 8 subjects rendered as cards, while one-pager mode renders exactly 1 subject as a detailed briefing. Each subject allows 2–6 key facts, 1–6 tags, a 2–5 sentence summary, and up to 5 activity items.

Four palettes are available: warm-paper (cream and terra-cotta, the default), deep-ink (dark with gold accent), clean-white (white with green accent), and slate (gray with blue accent). The palette is applied via a body class using CSS custom properties.

The interview runs only when the provided data or intent is incomplete. It collects subject type, count, purpose, key facts, narrative emphasis, and palette in one message so the resulting JSON document passes the validation checklist before rendering.