Use this skill when the user needs visual proof that specific values exist in a PDF — not just to read a PDF, but to see exactly where a number, amount, clause, or field appears on the page with a highlighted screenshot. Trigger when someone is cross-referencing a PDF against something else (a form they're filling out, a claim, a conversation, another document) and needs confirmation with evidence. Common signals: entering data into TurboTax or a form and pulling source numbers from last year's return; asking an accountant to verify reported figures; checking contract terms before sending to legal; matching invoice line items to a PO. The core intent is 'show me the actual text in context' — not summarize, not extract all text, but produce a cropped screenshot with the value highlighted. Do NOT trigger for PDF operations without a specific value to locate: merging, splitting, summarizing, converting, or creating PDFs.
PDF reading, creation, and review guidance
## Reading PDFs
- Use `pdftoppm -png $OUTDIR/$BASENAME.pdf $OUTDIR/$BASENAME` to convert PDFs to PNGs.
- Then open the PNGs and read the images.
- `pdfplumber
Split a PDF file into multiple smaller parts, each under a specified file size limit. Use this skill whenever the user wants to break a large PDF into chunks by file size — for example, to fit under an upload limit, email attachment cap, or API constraint. Trigger on phrases like "split PDF", "break PDF into parts", "PDF too large", "PDF under X MB", "chunk PDF", or any request to divide a PDF file by size. Always use this skill when a PDF splitting task involves a file size target, even if the user just says something like "my PDF is too big to upload".
[omh] Huge PDF or document to read in full: read a very large PDF, contract, manual, or report through Hermes in page-anchored ranges with a coverage ledger. Use when the user says: long-document-reading, long document reading, summarize this pdf, read this pdf, process this pdf, go through this pdf, summarize this document, read this document.
Convert text-layer or scanned PDF files to clean Markdown and extracted images with MinerU, then run deterministic quality checks and optional LLM-assisted formula review. Use when a user asks to OCR a PDF, extract a PDF to Markdown, preserve formulas or tables from a PDF, or prepare a PDF for translation.
PDF Render QA
Use for source or exported PDFs. Treat rendered pages as layout evidence; text extraction alone is insufficient.
## Stage: enhance
Preserve page, figure/table, caption, and crop provenance
PDF Visual Curation
Use only with PDF source attachments. It selects source evidence; ingest remains the source of figures and tables.
## Stage: enhance
Preserve the explicit image policy and distinguish
Use this skill any time a .pdf file is involved -- as input, output, or both. This includes: reading, parsing, or extracting text from a PDF; viewing outlines, stats, or formatting issues; rendering PDF pages to HTML or SVG; applying highlights, text color changes, or modifications on specific pages; deleting pages; replacing content stream text. Trigger whenever the user mentions 'PDF', 'pdf file', or references a .pdf filename.
Build fillable PDF forms and contract templates with signature fields. Use when creating a form or template, adding fillable fields to an existing PDF, replacing static text with fields, or preparing a document for e-signing. Triggers on "PDF form", "fillable PDF", "add fields to a PDF", "contract template", "skapa PDF-formulär", "PDF-mall", "fyllbart PDF", "formulario PDF". Not for sending a finished document: see formify-send-contract.
把 PDF 转成**可编辑**的 PowerPoint——文字变真文本框、色块和线条变原生矢量形状、 照片变图片，不是每页拍一张图。走 LibreOffice 的 PDF 导入过滤器，再自动修掉它那几个 对中文不友好的默认行为（字体名无空格 / 文本框叠字 / 汉字变康熙部首 / 逐字空格）。 当用户说「PDF 转 PPT」「pdf to pptx」「这份 PDF 我想改字」「只有 PDF 没有源文件」 「方案 PDF 转成可编辑 PPT」「把这个 PDF 做成 PowerPoint」时使用。 与 `html-to-pptx` 的区别：那个只吃 HTML，喂 PDF 会被直接拒。 扫描件（无文字层）不适用——只会得到一堆图片，得先 OCR。
name: react-pdf
description:
"Generate PDF documents using React-PDF library (@react-pdf/renderer). Use when creating PDFs,
generating documents, reports, invoices, forms, or when user mentions PDF generation, document
Convert a markdown file to PDF using mistune + reportlab. Use when the user wants to convert a .md file to PDF, or when another skill needs to produce a PDF from markdown output.
Translate English PDF documents (especially academic papers) into Chinese using PDFMathTranslate-next (pdf2zh), producing a monolingual Chinese PDF and a bilingual side-by-side PDF with formulas, figures, and layout preserved. Automatically installs the pdf2zh_next environment via uv and reuses the model/API the user already configured in Codex or Claude Code (any OpenAI-compatible gateway), falling back to the free SiliconFlow service when no key is present. Use this skill whenever the user wants to translate a PDF to Chinese (or another language), translate a paper/论文, produce a bilingual PDF, or mentions pdf2zh / PDFMathTranslate — even if they don't name the tool explicitly. Proactively invoke this skill (do not translate manually) whenever a PDF translation task appears.
Use whenever the user is customizing, overriding, or debugging PDF output in Perfex CRM — invoice PDFs, estimate PDFs, proposal PDFs, payment receipts, contract PDFs, statement PDFs, credit note PDFs, the `my_` prefix override convention, TCPDF library usage, `App_items_table` customization, font configuration (freesans, dejavusans, droidsansfallback), PDF merge fields, logo/heading settings, or e-invoice JSON/XML export (3.4.0+). Also trigger when the user says "my PDF is blank", "Arabic text broken in PDF", "custom logo not showing in invoice PDF", "how to add a field to the invoice PDF", "PDF font wrong", "override invoicepdf.php", "items table column in PDF", or "e-invoice XML format".
Convert PDF files to Markdown using opendataloader-pdf. Extracts text, tables, headings, lists, and images with correct reading order. Use for PDF parsing, PDF to Markdown conversion, document extraction, and AI-ready data preparation.
Plain text files in a repository that tell a coding agent how the project works: commands to run, conventions to follow and things to avoid. CLAUDE.md, AGENTS.md, cursor rules and skills are the common kinds.
CLAUDE.md or AGENTS.md?
CLAUDE.md is read by Claude Code. AGENTS.md is an open format that Codex, Cursor and other agents read. Many projects keep one and point the other at it.
What is a skill?
A folder with a SKILL.md that describes one capability, such as filling PDFs or reviewing code. The agent loads it only when the task calls for it.
Can I search my own team's files too?
Your agents already can, over MCP, limited to the files you're allowed to read. Searching them from this page is coming.
metedataSkill
