mcpField
F
filemorph

filemorph

Local
github.com/CloudCorpRecordsProductivity0Updated 1w ago

About filemorph

filemorph is a local MCP bridge and CLI for converting files through the hosted FileTypeConverters service. It exposes format lookup, conversion checks, submission, and retrieval without converting or storing files itself.

Key features

  1. 01Convert files across 44 formats
  2. 02Inspect formats and supported pairs
  3. 03Submit files as base64 or HTTPS URLs
  4. 04Retrieve earlier conversion results
  5. 05Use tools through a CLI

Use cases

  1. 01Check whether PDF-to-DOCX conversion is supported
  2. 02Convert an uploaded spreadsheet to CSV
  3. 03Fetch a previously submitted conversion
  4. 04Run file conversion commands from a terminal

MCP support

It exposes the upstream FileTypeConverters tools as a local MCP server and forwards calls to that hosted MCP server. Results are returned unchanged.

Configuration

Local server config is the same for every client. Copy it into your client’s MCP config file.

mcp.json · Universal config
{
"mcpServers": {
"cloudcorprecords-filemorph": {
"command": "npx",
"args": [
"-y",
"@reneturcios/filemorph"
]
}
}
}
Run commandnpx -y @reneturcios/filemorph

Local servers need a runtime on your machine (e.g. Node.js 18+). Config file locations differ by client; see your client’s docs.

Packages
npm@reneturcios/filemorph0.1.0
ocighcr.io/cloudcorprecords/filemorph:0.1.0

Tools

5

Saved tool list · updated 1w ago

This tool takes no parameters.

Details

From the project README

filemorph

filemorph is a small local bridge to the hosted FileTypeConverters MCP server, which converts files between 44 formats spanning documents, spreadsheets, images, ebooks and subtitles. You hand it a file either as base64 content or as a public https URL, pick a target format, and get the converted output back — small text outputs inline, larger or binary outputs as base64 plus a download URL.

The bridge converts nothing itself and keeps no files. It serves the upstream server's tool list locally, exactly as captured from the live endpoint, and forwards each tool call upstream, returning the result untouched. Two tools read reference data, one checks whether a conversion pair is supported, one submits a conversion, and one fetches a conversion submitted earlier.

Tools

  • list_formats — every format the converter can read or write, with its id, name, extensions and family. The ids are what the other tools take as from and to. No arguments.
  • check_conversion — whether one format can be converted to another, checked before any file is sent. Reports the engine, whether AI Deep Read is needed, the options the pair accepts, and alternatives when the pair is not supported. Formats may be given as ids (docx), extensions (jpeg) or common names (word, excel). Required: from, to.
  • list_conversions — the supported target formats for a source format, or every source that can produce a given target, as compact {from,to,slug,engine,mode} rows. Optional: from, to.
  • convert_file — submit a file for conversion. The content goes in as base64 or as a public https url; from is optional when the filename or URL carries an extension. Text outputs (Markdown, CSV, JSON, HTML…) up to 200 KB come back inline as text; other outputs come back as base64 plus a downloadUrl that accepts the same bearer token. Files are capped at 10 MB. Setting deepRead: true reads scans, photos and handwriting with AI Deep Read (1 credit per page). Required: to. Optional: filename, base64, url, from, options, deepRead.
  • get_conversion — a conversion submitted earlier, fetched by id: its status, any warnings, and the output (inline text or base64) while it is still retained. Required: id.

Configuration

  • FILETYPECONVERTERS_MCP_URL (optional) — override the upstream endpoint. Defaults to https://filetypeconverters.com/api/mcp. (The override hook is this rebuild's own labelled reconstruction.)
  • FILETYPECONVERTERS_API_KEY (optional) — when set, every upstream request carries an Authorization: Bearer <key> header. When it is not set, no auth header is sent. (This forwarding is also a labelled reconstruction; the original surface was captured keyless, though the registry record declares an Authorization header required.)
Forwarding mechanics

On the first tool call the client opens a streamable HTTP session with the upstream: a JSON-RPC initialize request (protocolVersion 2025-03-26, clientInfo name filemorph), then a notifications/initialized notification, then the tools/call request carrying the caller's arguments unchanged. When the upstream answers initialize with an Mcp-Session-Id header, that value rides on every later request of the session, which is reused for the life of the process. Replies may arrive as plain JSON or as server-sent events; both are read.

Successful upstream results pass through verbatim, including any isError flag the upstream sets. Failures — a JSON-RPC error, a non-2xx status, an unreadable response, or a request that cannot be completed — come back as an isError result whose text starts with filemorph error: (this rendering is this rebuild's own labelled reconstruction).

CLI

Each tool is also a subcommand taking one JSON argument string (omit it for tools that take none):

npx @reneturcios/filemorph list_formats
npx @reneturcios/filemorph check_conversion '{"from":"pdf","to":"docx"}'
npx @reneturcios/filemorph convert_file '{"to":"md","filename":"report.pdf","base64":"..."}'
npx @reneturcios/filemorph get_conversion '{"id":"conv-42"}'
npx @reneturcios/filemorph --version

Exit codes: 0 on success, 1 on an isError result, 2 on a usage error.

MCP server

npx -y @reneturcios/filemorph mcp

MCP client config:

{
  "mcpServers": {
    "filemorph": {
      "command": "npx",
      "args": ["-y", "@reneturcios/filemorph", "mcp"]
    }
  }
}

Docker

docker run -i ghcr.io/cloudcorprecords/filemorph
docker run --rm ghcr.io/cloudcorprecords/filemorph --version

The container's default command is mcp, so the first form is a stdio MCP server; any other arguments reach the CLI normally.

Development

npm install
npm run build
npm test
npm run dev

The test suite runs entirely against a stubbed JSON-RPC upstream — no live network calls.

License

MIT — see LICENSE.

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.