Spa to pdf

Skill Jim4546/ai-workflow-skills/spa-to-pdf

AI 工作流 Skill 合集：办公、研究、归档、日报、尽调、文档转换等可复用工作流

Install

npx -y skills add Jim4546/ai-workflow-skills --skill spa-to-pdf

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

2 things to look at

no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Scrape a single-page-app (SPA) website with multiple collapsible/sectioned content blocks and produce a Word document plus a PDF with an auto-filled table of contents and clickable inline hyperlinks. Use when the user provides a SPA URL (hash routes, JS-rendered content) and asks for a Word/PDF export of all sections. Requires Windows + Microsoft Word for the final PDF step.

SKILL.md

3.8 KB, 879 tokens by cl100k_base, as published. Nobody here has run it

spa-to-pdf

Turn a JS-rendered SPA page (e.g. https://example.com/#/some/section) into:

output.docx — cover, bilingual TOC, every section as Heading 1 with original paragraphs/lists/hyperlinks preserved
output.pdf — same, with TOC fields filled and heading bookmarks

When to invoke

Trigger this skill when the user asks to "scrape / save / export / archive a website's content into Word or PDF" AND the site looks like a SPA (URL has #/..., content is dynamic, or curl returns an empty shell). For plain static HTML pages, a simpler WebFetch + pandoc pipeline is enough — skip this skill there.

Prerequisites

Python 3.10+
Windows + Microsoft Word installed (needed for PDF + TOC fields)

One-time setup:

python -m venv .venv
.\.venv\Scripts\Activate.ps1
pip install playwright python-docx beautifulsoup4 lxml pywin32
playwright install chromium

Pipeline

Run the four scripts in scripts/ in order. They share a working dir (defaults to ./raw/ for intermediate files and ./output.docx, ./output.pdf for products). All scripts take CLI args — see --help on each.

1. Probe (optional but recommended)

python scripts/probe.py <URL>

Dumps the rendered HTML and the top CSS class names. Use the output to pick selectors for step 2. If the page is a typical Collapsible-based site (e.g. World Scholar's Cup themes), the defaults already work.

2. Scrape

python scripts/scrape.py <URL> `
    --section-selector .Collapsible `
    --title-selector .Collapsible__trigger `
    --content-selector .Collapsible__contentInner

Writes raw/index.json with {root_url, page_title, subtitle, sections: [{title, html, links}]}.

3. Build Word

python scripts/build_doc.py --in raw/index.json --out output.docx

4. Verify

python scripts/verify.py --in raw/index.json --docx output.docx

Exits 0 if every section's character count, hyperlink count, and word set roughly matches the source HTML. If not, the script prints what's missing per section. Always run this before showing the doc to the user.

5. Export PDF

python scripts/to_pdf.py output.docx output.pdf

Opens Word, updates TOC fields + bookmarks, exports PDF, closes Word.

Important behaviors

Inline links only. Do NOT visit each outbound link and append its content — the source page already places the links inline with context; duplicating the targets bloats the doc and adds noise.
No add_page_break(). Use paragraph_format.page_break_before = True on Heading 1 — manual page breaks cause trailing blank pages when content happens to fill a page.
TOC page title must NOT be Heading 1. Otherwise the TOC contains itself. Use a plain styled paragraph (see build_doc.py).
Always run verify.py before reporting success. It catches missing nested lists, dropped runs, and character-encoding bugs that aren't visible at a glance.

For the full list of edge cases and why each fix exists, read references/pitfalls.md. Read it before changing any of the scripts — several of the "obvious" simplifications will reintroduce bugs we already fixed.

Cross-platform note

The PDF step depends on pywin32 and Microsoft Word. On macOS this can be ported to AppleScript / appscript; on Linux, use LibreOffice headless with a TOC-update macro. PRs welcome.

Gives 0 of the 12 instructions most pdf office docs skills give in 879 tokens

Counted across 636 of the 690 authors here whose files we hold, read 2026-08-07

extract text using pdfplumberin 89 of 636, across 23 files
create PDFs using reportlabin 83 of 636, across 16 files
read forms.md to fill out pdf formsin 80 of 636, across 13 files
OCR scanned PDFs using pytesseractin 77 of 636, across 10 files
merge or split PDFs using qpdfin 70 of 636, across 3 files
use excel formulas instead of hardcoded calculated valuesin 68 of 636, across 13 files
unpack edit xml and repack existing documentsin 63 of 636, across 8 files
document sources for hardcoded valuesin 61 of 636, across 9 files
write minimal python code without unnecessary commentsin 59 of 636, across 7 files
run the recalculation script after adding or modifying formulasin 59 of 636, across 7 files
fix all identified formula errors and recalculatein 58 of 636, across 6 files
format years as text stringsin 57 of 636, across 5 files

Said here and by no other author read

skip this skill for plain static HTML pages
run the probe script to pick selectors
write scraped data to raw/index.json
build a Word document from the scraped JSON
run verify.py before reporting success
use paragraph_format.page_break_before for page breaks

Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.