Converters

Your agent can turn documents, email, and diagrams into Markdown, and turn Markdown into branded Word, PowerPoint, Excel, and styled HTML.

Install

You need agentbundle first. Install it once for your user account:

agentbundle install --pack converters

In Claude Code you can install it as a plugin instead:

claude plugin install converters@agent-ready-repo

What your agent can do

• Read a PDF, Office file, HTML page, EPUB, spreadsheet, or image and hand back Markdown your agent can use as context.
• Convert an Outlook or .eml email to Markdown with its headers and an attachments table.
• Fill your branded Word, PowerPoint, or Excel template from a Markdown file.
• Turn Markdown into a shareable, styled HTML page, or an offline proof page for review.
• Render Mermaid diagrams in a Markdown file as images.

Skills

• file-to-markdown — Converts documents and images to Markdown, with provenance and a quality signal in the output.
• msg-to-markdown — Converts Outlook .msg and .eml email to Markdown, keeping headers, body, and attachments.
• markdown-to-docx — Fills a branded Word template from a Markdown file.
• markdown-to-pptx — Fills a branded PowerPoint template from a Markdown file.
• markdown-to-xlsx — Fills a branded Excel template from a Markdown file.
• markdown-to-html — Makes a self-contained, styled HTML page with navigation, highlighted code, and diagrams.
• render-proof — Renders a Markdown draft as an offline HTML proof for human review.
• mermaid-renderer — Replaces each Mermaid block in a Markdown file with a rendered PNG or SVG.

Works with

Works with Claude Code, Codex, Cursor, GitHub Copilot, Gemini CLI, and Kiro IDE. Architect draws Mermaid diagrams that mermaid-renderer can turn into images. Guides: Convert documents to Markdown, Publish Markdown to Office, and Convert Markdown to HTML and email.

Questions

Do I need machine-learning tools installed? Not for common documents. file-to-markdown reads PDF, Office, HTML, EPUB, CSV, and email with plain parsers. Only .xls files and images need more.

Does it need my own templates? The Word, PowerPoint, and Excel skills fill templates your designer prepared. They do not invent a layout.

What if a library is missing? Each Office skill checks at run time and stops with the exact pip install line to run.

How does it read an image? It tiles the image, reads each tile with the model’s vision, and merges the tiles by script.

Details

Scope: user (default), repo · Version: 0.9.9 · Source: packs/converters