mk:wiki-render

What This Skill Does

Renders a wiki slug into a single self-contained HTML snapshot — every page with its provenance, openable offline.

When to Use

• Triggers: share or review accumulated wiki knowledge as a static page.
• NOT for: capturing/approving knowledge (mk:wiki) or generic code/diagram/diff visuals (mk:preview).

Command

npx mewkit wiki render <slug>              # print HTML to stdout
npx mewkit wiki render <slug> --out <path> # write HTML to a file

Guarantees

• Self-contained / no CDN — inline CSS only, zero http(s):// asset references, no <script>, no @import. Opens with no network.
• DATA-safe — page content is HTML-escaped before embedding, so wiki content (which is DATA) cannot inject markup or script into the rendered artifact.
• Provenance shown — origin, approver, and source ids render as escaped text (never as live links).

Gotchas

• Render reads the canonical files under tasks/wikis/<slug>/; mewkit wiki reindex is NOT required (render reads files directly, not the DB).
• A source URL in provenance appears as escaped TEXT, not a clickable <a href> — by design (no remote/clickable assets in the offline snapshot).
• Empty wiki → a valid HTML page with "No pages yet."

Peer Skills

mk:wiki (capture + approve) · mk:wiki-research (external fetch → candidate)