菜单
CLI for converting files to Markdown using Microsoft's markitdown. Use when converting PDF, DOCX, PPTX, XLSX, HTML, images, audio, or other file formats to Markdown for text analysis or LLM processing. Triggered by requests involving file-to-markdown conversion, document text extraction, or preparing files for LLM input.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
GitLab CLI for merge requests, issues, CI/CD pipelines, repositories, and more. Use when working with GitLab to list/view MRs, check CI status, browse issues, or any GitLab API operation. Triggered by requests involving GitLab data, merge requests, CI pipelines, or GitLab project management.
Judgment rules for how to shape commits, plan branches, and respond to review — not git mechanics. Use when staging changes, writing commit messages, splitting/reordering commits, planning a branch's structure, addressing review feedback, or deciding how to merge.
This skill should be used when the user asks to "search the web", "look up online", "find recent information", "search for", "what's the latest on", or needs web search results. Provides access to a private SearXNG metasearch engine instance via the `searxng` CLI. Prefer this over WebSearch when available.
CLI for querying RabbitMQ instances (vhosts, queues, exchanges, bindings). Use when inspecting RabbitMQ, listing queues, checking exchange bindings, viewing queue details, or exploring RabbitMQ topology. Triggered by requests involving RabbitMQ data, queue inspection, exchange listing, or message broker exploration.
Use Gemini CLI in non-interactive mode for tasks requiring massive context windows (1M tokens). Best delegated to subagents for iterative analysis and summarization. Invoke when analyzing large codebases, requesting deep analysis, getting second opinions on complex problems, or when Claude's context limits are insufficient. Triggers include phrases like "use gemini", "analyze with gemini", "get second opinion", "deep analysis of codebase", or when processing files exceeding Claude's context capacity. IMPORTANT - Always delegate to a subagent using the Task tool for better iteration and result summarization.
CLI for querying Metabase (cards/questions, dashboards, collections, databases, search). Use when working with Metabase to search entities, browse collections, run saved questions, export/import dashboards, explore database metadata, or resolve Metabase URLs. Triggered by requests involving Metabase data, dashboard queries, saved questions, collection browsing, or Metabase automation.
基于 SOC 职业分类
| name | markitdown |
| description | CLI for converting files to Markdown using Microsoft's markitdown. Use when converting PDF, DOCX, PPTX, XLSX, HTML, images, audio, or other file formats to Markdown for text analysis or LLM processing. Triggered by requests involving file-to-markdown conversion, document text extraction, or preparing files for LLM input. |
| trigger-keywords | markitdown, convert to markdown, file to markdown, extract text, pdf to markdown, docx to markdown, pptx to markdown, xlsx to markdown |
| allowed-tools | Bash(markitdown:*) |
Command-line tool for converting various file formats to Markdown. Built by Microsoft for LLM and text analysis pipelines.
markitdown document.pdf # Convert, output to stdout
markitdown document.pdf -o output.md # Convert, write to file
cat document.pdf | markitdown # Convert from stdin
markitdown document.pdf > output.md # Redirect stdout to file
markitdown < document.pdf # Redirect stdin
Converted without any extras: HTML, CSV, JSON, XML, plain text, ZIP (iterates contents), EPub, Jupyter notebooks (.ipynb), RSS feeds, Wikipedia URLs, Bing SERP results.
With extras installed: PDF, DOCX, PPTX, XLSX, XLS, Outlook MSG, images (EXIF metadata + optional LLM captioning), audio (EXIF metadata + optional transcription), YouTube URLs (transcript extraction).
markitdown -v # Show version
markitdown -o FILE # Write output to file instead of stdout
markitdown -x .pdf # Hint file extension (useful with stdin)
markitdown -m application/pdf # Hint MIME type
markitdown -c UTF-8 # Hint charset
markitdown --keep-data-uris # Keep base64-encoded data URIs in output (truncated by default)
For higher-quality extraction using cloud OCR:
markitdown -d -e "https://your-endpoint.cognitiveservices.azure.com" document.pdf
Third-party plugins extend format support via the markitdown.plugin entry point:
markitdown --list-plugins # List installed plugins
markitdown -p document.xyz # Enable plugins for conversion
markitdown report.pdf | wc -w # Word count
markitdown report.pdf | head -50 # Preview first 50 lines
for f in *.pdf; do markitdown "$f" -o "${f%.pdf}.md"; done
curl -sL "https://example.com/doc.pdf" | markitdown -x .pdf
When piping content, provide an extension hint so markitdown can select the right converter:
cat unknown_file | markitdown -x .docx