Hushvert MCP

@hushvert/mcp

A file-conversion tool for your AI agent.

@hushvert/mcp is a Model Context Protocol server that gives an AI coding agent (Claude Code, Cursor, Cline, Zed, and any other MCP host) a convert_file tool over the hushvert hosted API. When a task needs a conversion a browser cannot do - office documents to PDF, PDF to Word, large video transcodes - the agent converts the file in one tool call, and the result is written next to the input. No upload code, no polling, no glue.

It is a thin client over the hushvert hosted API. For the conversions that DO run in a browser (images, HEIC, archives, audio, small video, PDF page ops), use the free, open-source @hushvert/engine package instead - this server will refuse those and point you there.

Demo

A real run in Claude Code, recorded live and not sped up: ask, and report.pdf is written next to the input. The whole turn took 23 seconds, of which the conversion itself was about 7.

The recording starts Claude Code with only this server loaded (--strict-mcp-config) so nothing unrelated is on screen. The tape that produced it is assets/demo.tape, if you want to reproduce it.

Claude Code can already convert files. Why this?

Because it can only convert what your machine can convert, and when it cannot, it does not fail loudly.

Ask any coding agent to turn report.docx into a PDF. If LibreOffice is installed, it will shell out to soffice, do a good job, and you do not need this server. If LibreOffice is not installed, and it is not there by default on macOS, on Windows, or in a typical CI image, the usual fallback is pandoc. Pandoc does not really convert a Word document. It reads the text into its own AST, hands that to LaTeX, and LaTeX typesets a new document. You get a PDF. The agent reports success. Nobody opens the file.

Here is the same report.docx down both paths:

Same words, different document. Every font in the pandoc PDF is Latin Modern, LaTeX's default. The heading color is gone, the table lost its Word styling, and the title moved into a centered LaTeX title block. pdffonts on the two outputs:

pandoc     LMRoman17-Regular, LMRoman12-Bold, LMRoman10-Italic, ...

hushvert Carlito-Regular, Carlito-Bold, Carlito-Italic

Carlito is metric-compatible with Calibri, which is what the document actually asked for. Latin Modern is not.

So, honestly:

• If LibreOffice is installed and your agent reaches for it, you do not need this server. That is a real answer, and it is the right one for a lot of people.
• If it is not installed, this is one line of config instead of a 281 MB download, and it behaves the same on your laptop, in CI, in a container, and on a machine you are not allowed to install software on.
• Name the tool if it matters. On a machine that had both this server and pandoc available, we asked the plain way ("convert report.docx to PDF") twice: Claude Code used convert_file once and pandoc the other time, and the two runs produced the two documents above. Which tool an agent reaches for is its call, not ours. "Convert report.docx to PDF with hushvert" pins it.

Install

Get an API key at hushvert.com/developers/keys (sign-in is a one-time email code; keys require a confirmed email), then add the server to your agent's MCP config.

Claude Code - one line, no file to edit:

claude mcp add hushvert -e HUSHVERT_API_KEY=hv_live_your_key_here -- npx -y @hushvert/mcp

Cursor, Cline, Zed, and other MCP hosts (or Claude Code, if you prefer a committed project config) - add this block to the host's MCP config (.mcp.json, .cursor/mcp.json, and so on):

{

"mcpServers": { "hushvert": { "command": "npx", "args": ["-y", "@hushvert/mcp"], "env": { "HUSHVERT_API_KEY": "hv_live_your_key_here" } } } }

If that file is committed, do not put the key in it. Claude Code expands environment variables in an MCP config, so use "HUSHVERT_API_KEY": "${HUSHVERT_API_KEY}" and keep the real key in your shell. Other hosts vary; check yours before committing.

Then ask your agent: "convert report.docx to PDF with hushvert." Naming the server is worth the two extra words: if your machine has a local converter, the agent may reach for that instead, and for office documents the result is usually worse. See above.

Tools

convert_file

input_path       (required) path to the source file

to (required) target format, e.g. "pdf", "docx", "mp4" from (optional) source format; inferred from the extension otherwise output_path (optional) where to write; defaults beside the input. Required to overwrite. wait_seconds (optional) max seconds to wait before handing back a jobId to poll. Default 120. idempotency_key (optional) makes a retried conversion safe (same job, charged once)

Returns { output_path, jobId, pair, bytesIn, bytesOut, status }. If the job is still running after wait_seconds (typical for large video), it returns { jobId, status: "processing", resumeWith: "convert_poll" }; call convert_poll with the jobId and an output_path to finish.

What it converts

The server-only formats a browser cannot do:

• Office to PDF: docx, pptx, xlsx, doc, ppt, xls, odt, ods, odp, rtf, html to pdf
• PDF to Word: pdf to docx
• Document interchange: md, html, epub, latex, rst, docx (via pandoc)
• Video: mov, mkv, avi, webm to mp4 (and mp4 to gif)

Call list_formats for the live list. Everything else (images, HEIC, archives, audio, small video, PDF page ops) runs free, client-side, in @hushvert/engine.

Configuration

Billing and privacy

Conversions are billed per use against your account: a free monthly allowance, then credits. check_usage shows your remaining allowance and balance at any time - have your agent check it before a large batch. The hosted API processes the server-only formats that genuinely cannot run in a browser; for everything else the file never leaves the device via the open-source engine. See hushvert.com/for-developers.

Security

The API key is read from the environment, never logged, and never returned in a tool result or error. See SECURITY.md.

License

MIT