)` throughout (C26): the inline\nfallbacks make a fresh app render correctly with zero setup, and a host brands\nit by passing `tokens: \":root{--matrx-print-accent:#0f766e}\"` — appended last,\nso the host wins. Both sheets carry the print hard parts (`@media print`,\npage-break avoidance, orphans/widows, `print-color-adjust`).\n\n### `@ai-matrx/print/directives` — page breaks, one grammar everywhere\n\n```ts\nimport { isPageBreakLine, splitAtPageBreaks, PAGE_BREAK_MARKDOWN } from \"@ai-matrx/print/directives\";\n```\n\nPut `` on its own line to start a new printed page\n(`\\pagebreak`, `\\newpage`, `` and\n`` are accepted too). Both\nconverters render it as `` — a dashed \"Page\nbreak\" rule on screen, a real page break on paper — and `markdownToPdfBlob`\nstarts a new PDF page there. The entry has no dependencies, so a chat or editor\nrenderer can recognise the same syntax without loading the stylesheets.\n\n### `@ai-matrx/print/pdf` — DOM capture & markdown → PDF\n\n```ts\nimport { captureToPDF, captureElementsToPDF, markdownToPdfBlob } from \"@ai-matrx/print/pdf\";\n\nawait captureToPDF(document.getElementById(\"report\")!, { filename: \"report\", orientation: \"landscape\" });\nconst blob = await markdownToPdfBlob(markdown); // converter + stylesheet included\nconst branded = await markdownToPdfBlob(markdown, { skin: \"article\", tokens }); // tune without replacing\nconst custom = await markdownToPdfBlob(markdown, { convertToHtml, loadCss }); // override either port\n```\n\n`markdownToPdfBlob` needs nothing from you. Both ports ship a working default\n(they were REQUIRED injections until 0.3.0); supplying one replaces that\ndefault and leaves the other alone. The defaults load through a lazy\n`import(\"@ai-matrx/print/markdown\")`, so a consumer that only does DOM capture\nnever pulls in the stylesheets.\n\nNothing degrades silently: a flashcard missing its `Front:`/`Back:`, markdown\nthat converts to an empty document, an image that 404s or never finishes\nloading (html2canvas would paint a blank rectangle and say nothing), and a\nfailed rasterization all report through `onNotice` — or `console.warn` with the\nremedy when no sink is given. `imageTimeoutMs` (default 8000) bounds the wait.\n\nhtml2canvas + jsPDF, lazy-loaded on use. Handles the Tailwind v4 problem\n(oklch/lab colors html2canvas can't parse) via a scoped `getComputedStyle`\npatch that is always restored. `captureElementsToPDF` renders deliberately\npaginated report pages one-per-PDF-page; the single-element lane slices one\ntall canvas across pages.\n\n### `@ai-matrx/print/react` — the options dialog & sheet preview\n\n```ts\nimport { PrintOptionsDialog, usePrintOptions, LabelSheetPreview } from \"@ai-matrx/print/react\";\n```\n\n`PrintOptionsDialog` renders any `BlockPrinter`'s variants + settings\n(desktop dialog, mobile bottom drawer) and runs the print.\n`LabelSheetPreview` shows a scaled, exact-proportion preview of a label sheet,\nwith a calibration outline mode. Tailwind CSS classes; Radix dialog + vaul\ndrawer are bundled dependencies.\n\n## License\n\nMIT\n\n### `@ai-matrx/print/document` — print-grade documents (PDF, Word, EPUB, HTML)\n\n```ts\nimport { exportDocument, downloadDocumentExport } from \"@ai-matrx/print/document\";\n\nconst exp = await exportDocument(markdown, \"docx\"); // \"pdf\" | \"docx\" | \"epub\" | \"html\" | \"markdown\"\ndownloadDocumentExport(exp); // browser: saves \"
.docx\"\n```\n\nSettings ride in the markdown's frontmatter:\n\n```yaml\n---\ntitle: Quarterly Review\nsubtitle: Operations\nauthor: Jane Park\ndate: today\ncover: true\ntoc: true # or { depth: 2, title: Contents }\npage: { size: a4, orientation: portrait, margins: 1in }\nheader: \"{title} | | {date}\" # left | center | right\nfooter: \"Page {page} of {pages}\"\ncolumns: 1\nnumberSections: true\ncsl: apa # apa | mla | chicago\nreferences:\n - { id: smith2020, title: Paper Cuts, author: \"Smith, Jane\", issued: 2020, publisher: Acme }\n---\n```\n\nIn the body: `<!-- pagebreak -->`, `<!-- section landscape columns=2 -->`,\n`[[toc]]`, `{#fig:chart}`, `Table: Caption {#tbl:sales}`,\n`## Methods {#sec:methods}`, `see @fig:chart`, `[@smith2020, p. 4]`.\nAll four formats come from one parsed tree (`parseDocument`), so they agree.\n\nCitations are formatted by **citeproc-js** over the official **CSL styles**:\n`csl:` accepts any style name from the CSL repository (zotero.org/styles) or a\n`.csl` URL; `apa`, `mla`, `chicago`, `chicago-notes`, `ieee` and `harvard` are\nbundled (offline). References come from the frontmatter and from your app via\n`exportDocument(md, fmt, { resolveReferences: async (keys) => cslItems })`.\nPDFs are hyphenated with Knuth–Liang patterns for the document's `lang`, and\nemoji / symbols print through Noto Emoji and Noto Sans Math.\n\nCSL styles and locales © the Citation Style Language project, CC BY-SA 3.0.\nNoto Emoji and Noto Sans Math are SIL Open Font License fonts (loaded at\nexport time from jsDelivr).\n`DocumentPrintPreview` (in `/react`) shows the real PDF pages with Print and\nevery download.\n\n","readmeFilename":"README.md"}