Word, Excel, PowerPoint, PDF, meeting notes, and manuals (3.6K skills)
Create, edit, and analyze spreadsheet files such as Excel workbooks and CSVs. Work with formulas and formatting, preserve existing formulas when making changes, analyze data, build visualizations, and recalculate formulas in supported workbook workflows.
Use this skill to read text and tables from PDFs, create documents, combine or split files, and handle forms. It gives an agent practical workflows for inspecting and processing PDF documents, including scanned files and batch tasks.
Use this skill to create, edit, or inspect PowerPoint presentations. It supports working with slide layouts, text, comments, and speaker notes, with workflows for starting from scratch, adapting a template, or changing an existing deck.
Create, edit, and analyze professional DOCX files. The workflow covers preserving formatting, extracting text, adding comments, and working with tracked changes, with guidance for generating new documents and editing existing ones.
Use this skill to write or review technical design documents with a consistent structure. It covers problem context, goals, architecture and flows, concurrency, errors, deployment, and unit tests, with guidance for diagrams, test cases, and design constraints.
Create Mermaid diagrams that render clearly in GitHub Markdown. The guide covers supported diagram types, simple syntax, readable layouts, and basic mind maps, plus previewing diagrams and reviewing automated edits for rendering issues.
Write, review, or reorganize Windows repository docs and agent instructions around the reader’s task. Check claims against current source, preserve facts during restructuring, validate with existing documentation checks, and report unresolved claims and work performed.
Create, read, edit, and review Word DOCX documents, including comments and tracked changes. The workflow preserves document structure, checks page rendering, and supports detailed work with tables, fields, templates, and the underlying OOXML format used inside DOCX files.
Create, inspect, and edit Excel workbooks while checking formulas, formatting, links, macros, and other workbook content. The workflow recalculates formulas, validates the result, and renders sheets for visual review, with explicit handling for data that cannot be safely preserved.
Create or edit PowerPoint decks and templates, then validate and render them for review. It helps preserve slide structure, notes, comments, media, and user metadata through delivery.
Read, create, and review PDFs with checks of both content and appearance. Uses Python tools for document generation and text extraction, then inspects rendered page images for spacing, alignment, and readability.
Create a durable, beginner-friendly Markdown guide to an entire code repository. It maps the project structure, studies the source, and writes an overview plus numbered deep dives with real code excerpts explaining architecture, data models, subsystems, and how components fit together.
Translate explanatory prose in codebase documentation while preserving code, filenames, identifiers, paths, and technical terms in English. It can produce a parallel translated folder or handle one file bilingually or as a sidecar copy, defaulting to Chinese when no language is specified.
Use this skill to write and maintain `lat.md/` documentation that explains a project's architecture, design decisions, domain rules, and test specifications. It guides section structure, links between docs and code, and validation of references.
Translate text between languages by identifying the source and target languages, then presenting both the original and translated text. When the provided translation script is available, the skill uses it; the material does not specify particular languages or translation features.
Create and edit Obsidian notes using its Markdown conventions, including wikilinks, embeds, callouts, and YAML properties. The guide also suggests ways to connect notes and organize a vault for easier browsing and search.
Retrieve a named reference document for a skill when its main description is not enough. Use it to look up detailed technical guidance, examples, or advanced usage; the reference name is optional and defaults to `REFERENCE.md`.
Create, read, clean, edit, analyze, or convert spreadsheet files such as XLSX, XLSM, CSV, and TSV. The skill guides formula-based calculations, formatting, and error checks, with a spreadsheet file as the deliverable.
Capture the visible content of the current browser page as an image, with an optional save path. Use it to show page state, document results, or help debug a webpage. Desktop and other application screenshots are outside its scope.
Create a PowerPoint presentation draft from a topic or outline. It can suggest an outline, organize the content, and apply a polished layout using available templates. A Baidu API key is required for the generation workflow.
Parse document content with PaddleOCR-VL-1.5, including document structure, tables, formulas, and mixed text and images. It supports multilingual documents and requires a Baidu API key.
Create and assess presentations by organizing the message with the Pyramid Principle: lead with a conclusion, then support it with grouped arguments and evidence. The skill can produce PPTX or Markdown slides, charts, speaker notes, and a scored quality review.
Create, read, and update PowerPoint .pptx files, including pitch decks and presentations. Extract slide text, work with templates, layouts, speaker notes, or comments, and combine or split presentation files as needed.
Extracts text from images using the PP-OCRv5 lightweight model through PaddleOCR. It supports text detection and recognition in natural scenes and complex documents, including multilingual text, with scripts for general, accurate, and handwriting recognition.