Connect Claude to MarkItDown

Convert web pages and documents - PDF, Word, Excel, PowerPoint, HTML, CSV, EPUB and more - to Markdown an agent can read. Toolspoke puts 3 of its actions behind one MCP endpoint that Claude, Cursor, and Codex all speak.

Connect View pricing

Connection: Sandboxed CLI
Authentication: None
Actions exposed: 3
What a call bills: Sandbox time only
Adapter: Maintained by Toolspoke

Connected in three steps

1
Install MarkItDown
Open the marketplace in your workspace, add MarkItDown to the project your agents work in, and it appears on the gateway immediately.
2
Connect the credential
Nothing to connect. MarkItDown needs no credential.
3
Point your agent at the gateway
Give your client one address, https://toolspoke.com/mcp. Claude Code takes it as a command, Claude and Claude Desktop add it as a custom connector, and Cursor, Codex and VS Code each read it from a config file of their own.

.mcp.json

{
  "mcpServers": {
    "toolspoke": {
      "type": "http",
      "url": "https://toolspoke.com/mcp"
    }
  }
}

One block covers every tool you have installed. MarkItDown shows up in the client as soon as your policy allows it, and so does everything else you install later.

Where the address goes, per client

Claude Code: Run it in your project, then /mcp to sign in
claude mcp add --transport http toolspoke https://toolspoke.com/mcp
Claude and Claude Desktop: Settings, then Connectors, then Add custom connector
https://toolspoke.com/mcp
Cursor: ~/.cursor/mcp.json, or .cursor/mcp.json for one project
{ "mcpServers": { "toolspoke": { "url": "https://toolspoke.com/mcp" } } }
Codex: ~/.codex/config.toml
[mcp_servers.toolspoke] url = "https://toolspoke.com/mcp"
VS Code: .vscode/mcp.json, or the MCP: Add Server command
{ "servers": { "toolspoke": { "type": "http", "url": "https://toolspoke.com/mcp" } } }

What MarkItDown asks for

MarkItDown needs no credential. Install it and it works, and there is nothing of yours for Toolspoke to hold.

What Claude can do in MarkItDown

3 actions, each one declared and named by the connector rather than discovered at runtime. A workspace policy grants a person all of them, a hand-picked selection, everything on the read side, everything on the write side, or none.

Reads: 3Reads
Writes: 0Writes
Destructive: 0Destructive

Reads

Fetches data and changes nothing.

version
Report the installed MarkItDown version. Needs no network and no input, which is why it is the health check.
convert_url
Fetch a web page or a document at a public URL and return it as Markdown. Works for HTML pages, Wikipedia articles, YouTube videos (title, description and transcript when one exists), RSS feeds, and links straight to a PDF, DOCX, XLSX or PPTX. The sandbox reaches the public internet only; private and loopback addresses are not routable. Output is truncated at 100,000 characters.
convert_file
Convert a file that is already in the project's sandbox workspace to Markdown - PDF, DOCX, PPTX, XLSX, XLS, CSV, HTML, JSON, XML, EPUB, Jupyter notebook, ZIP (each member converted) or Outlook MSG. Relative paths resolve against /workspace. The file has to be there first, for example written by the Bash connector or downloaded with curl. Output is truncated at 100,000 characters.

What it will not do

Enforced by the gateway rather than left to convention, which is why each of these can be stated flatly.

It cannot call anything else: The 3 actions above are the whole of it. A call to any other name is refused before it reaches MarkItDown rather than forwarded on, and connecting your account does not add to the list: it is fixed by the connector, not discovered at run time.
It only reads: Every action here reads. Nothing this connector can do changes anything in MarkItDown.
It reaches no further than your credential: Toolspoke holds no access to MarkItDown of its own. Every call carries the credential you stored and nothing besides, so whatever that credential cannot reach, this connector cannot reach either.
It never hears from MarkItDown: Nothing is pushed to it. There is no webhook, no subscription and no polling, so this connector cannot notice by itself that something changed in MarkItDown. An agent has to ask.
It does not smooth over provider limits: Toolspoke does not retry, queue or back off around MarkItDown's own rate limits. A call that MarkItDown refuses comes back to the agent as a failed call.

Before you connect it

What can Claude do in MarkItDown?

3 named actions: 3 that only read. They include version, convert_url and convert_file. Nothing outside that list is reachable: the connector declares each operation by name rather than proxying whatever an agent asks for.

What credentials does the MarkItDown connector need?

None at all. MarkItDown is reachable without one, so there is nothing of yours for Toolspoke to store.

Does the MarkItDown connector work with Cursor and Codex, or only Claude?

Any client that speaks MCP, and every one of them gets the same 3 actions. There is a single address, https://toolspoke.com/mcp. Claude Code adds it with claude mcp add --transport http, Claude and Claude Desktop take it as a custom connector in settings, Cursor reads it from .cursor/mcp.json, Codex from ~/.codex/config.toml, and VS Code from .vscode/mcp.json. Each of them signs in to the gateway itself, so there is no key to paste.

What does the MarkItDown connector not do?

The 3 actions above are the whole of it. A call to any other name is refused before it reaches MarkItDown rather than forwarded on, and connecting your account does not add to the list: it is fixed by the connector, not discovered at run time. Every action here reads. Nothing this connector can do changes anything in MarkItDown. Toolspoke holds no access to MarkItDown of its own. Every call carries the credential you stored and nothing besides, so whatever that credential cannot reach, this connector cannot reach either. Nothing is pushed to it. There is no webhook, no subscription and no polling, so this connector cannot notice by itself that something changed in MarkItDown. An agent has to ask. Toolspoke does not retry, queue or back off around MarkItDown's own rate limits. A call that MarkItDown refuses comes back to the agent as a failed call.

Can I limit which actions an agent can call?

Yes, in two places. The project switches MarkItDown's actions on and off one at a time, for everyone in the project at once, and the screen groups them by read, write and destructive so turning off everything that deletes is one click. An individual agent key can then be narrowed further, to particular toolkits in a project and to particular actions in a toolkit. Whatever it was granted, a key never reaches a project its owner cannot.

What gets recorded when an agent calls MarkItDown?

Every attempt, with the agent that made it and the person that agent belongs to, the full request payload, the response payload, the status, the duration, and what the call cost in sandbox time and model tokens. Values whose key names a secret are masked out before the record is shown to anyone. An operation the connector marks as not retained never has its response body written at all, so the gateway keeps no second copy of what was read.

More data connectors

Browse all of them