Tools

From Microsoft: converts PDF, Word, Excel, PowerPoint, HTML and other files or URLs to Markdown so agents can read and search them. A single convert_to_markdown tool over STDIO or local HTTP.

Anthropic's document skills, the implementation behind Claude's file features: create, edit and analyze Word, Excel, PowerPoint and PDF files, including tracked changes, comments, formulas and form filling. Source-available, not under an open-source license.

Extracts text and tables from PDFs; merges, splits and rotates pages, adds watermarks, fills forms and OCRs scanned documents. The official Anthropic skill behind Claude's document features; source-available, not open source.

OCRs photos of visa documents such as bank, employment or income certificates, translates them into English and builds a side-by-side bilingual PDF. Needs Python (pillow, reportlab) and an OCR engine. Not an official certified translation: verify names, amounts and dates.

Reads messy invoices and receipts (PDFs, scans, screenshots), extracts vendor, date and amount, renames and files them by vendor, category or year, keeps originals and can export a CSV for your accountant. Needs an agent with local file access; no account required.

Modern markup-based typesetting system, a LaTeX alternative for papers, reports and resumes. Fast incremental compilation; typst compile outputs a PDF in one command and watch mode recompiles automatically.

Translates full PDF papers while preserving formulas, charts and layout (command pdf2zh), outputting monolingual and side-by-side bilingual PDFs. Uses Google Translate by default and can switch to DeepL, OpenAI and others; also runs as an MCP server.

Adds a searchable, copyable text layer to scanned PDFs, with automatic deskew, rotation and size optimization, outputting PDF/A by default. Built on Tesseract, supports 100+ languages and runs fully locally.

Anthropic's official small-business skill for bills: reads vendor, amount and due date from email, PDFs or photos, dedupes, codes to accounts, matches POs and drafts a payment plan. Posting and paying each need approval; it never auto-pays. Works with QuickBooks, Xero, NetSuite or file export.

Open-source e-book manager whose command-line tools (ebook-convert, ebook-meta, calibredb) batch-convert between EPUB, MOBI, AZW3, PDF, DOCX and more and manage library metadata. The install command is Linux-only; on macOS and Windows use the installer.

From Firecrawl: converts Word, PowerPoint, Excel, PDF, EPUB, RTF and more into clean Markdown locally in milliseconds, keeping headings, tables, footnotes and formulas so agents can read attachments. Needs Node.js; optional Firecrawl cloud OCR for scans uploads the whole file.

Document layout skill by tw93 with eight templates (resume, one-pager, long doc, formal letter, portfolio, slides, research report and more) that render polished PDF, PNG or editable PPT and check layout and content gaps. Supports English, Chinese, Japanese and Korean. Needs Python (WeasyPrint).

Optimizes websites for AI search: checks whether ChatGPT, Perplexity and Google AI Overviews can cite your site, covering AI crawler access, llms.txt, structured data, E-E-A-T and technical SEO, and writes client reports. PDF reports and scoring need the README install script, Python 3.8+ and Git.

Reads the exact position of every character, line and table cell in digital PDFs: the CLI exports layout-preserving text or CSV/JSON details, and the Python API extracts tables and form fields, ideal for invoices, statements and reports. Runs locally with Python; scans need OCR first.

PDF translation CLI for research papers that preserves formulas and layout as much as possible, producing bilingual side-by-side or translation-only PDFs. Needs an OpenAI-compatible API (local Ollama works too); currently focused on English to Chinese.

PDF command line written in pure Go with almost no external dependencies: dozens of subcommands to merge, split, crop, rotate, watermark, encrypt and decrypt, extract images and fonts, and work with forms.

Open-source alternative to NotebookLM's podcast feature: turns web pages, PDFs, images, YouTube videos or text into two-host conversational audio, useful for lecture notes or articles. Uses Gemini and OpenAI API keys by default, or local models plus free Edge TTS.

Self-contained modern TeX/LaTeX engine based on XeTeX and TeX Live that downloads and caches support files on first build, compiling .tex to PDF in one command without installing a full TeX distribution.

Open-source PDF table extraction with a camelot CLI: exports tables from financial reports, bank statements and statistical reports to Excel, CSV or Markdown, with an accuracy score. Needs Python 3.10+; handles text-based PDFs by default, scans need the ml and ocr extras. Double-check results.

PDFMathTranslate 2.x (command pdf2zh_next), built on BabelDOC: translates papers and other PDFs while keeping formulas, charts, TOC and annotations, outputting translated and bilingual side-by-side versions. Uses a free translation channel by default, or your own OpenAI, DeepSeek or other API key.

Converts PDF, Word, Excel, PowerPoint, images, audio, web pages and YouTube transcripts into Markdown that agents can read and cite. Built on Microsoft MarkItDown; the Docker image lacks audio transcription and image recognition, so build locally for full features.

Extracts issuer, amount, date, invoice number and other fields from PDF invoices using YAML/JSON templates, exporting CSV, JSON or XML. Ships several text-extraction backends, OCR for scanned invoices and an optional AI fallback.

Translates entire books: splits PDF, DOCX or EPUB into chunks for parallel subagents, keeps terminology consistent, resumes interrupted runs and verifies with hashes, then outputs HTML, DOCX, EPUB or PDF. Needs Calibre, Pandoc and Python.

Local text-to-speech CLI built on the open Kokoro model: batch-converts TXT, EPUB and PDF to MP3 or WAV by chapter, with multiple languages, voice blending and speed control; handy for audio versions of course materials. Needs Python 3.11-3.12 and two model files downloaded first.