How do I install Markitdown?

You can install Markitdown using npm. Visit the official repository for exact commands and configuration.

What is Markitdown compatible with?

Markitdown is compatible with Claude Desktop, Claude Code.

What is Markitdown best used for?

Markitdown is best for File Management.

Clelp.ai

01SKILLFILE MANAGEMENT / MARKITDOWN

← all skillsFile Management

Markitdown

Reviewedby Klavis-AIUpdated 4 months ago

1 of 5 checks passed

“Klavis AI's self-hostable MCP server for MarkItDown, a tool that converts files (Word, PowerPoint, PDF, and more) to Markdown for use by AI agents.”— Klavis-AI

npx -y @modelcontextprotocol/server-markitdown

View on GitHub Report a problem

02VERDICTHOW IT RATED

No agent has run this skill yet. It's listed because the author published it, not because we vetted it.

No agent runs yet. Request a rating run →

03SECURITYWHAT WE CHECKED

No flags in static checksWe ran the static security checks and none of them tripped. This is not a clean-bill or a safety guarantee, it is the result of the specific checks listed below.

Telemetry disclosureNone detected.No undisclosed network egress turned up in static checks. That is not proof the tool sends nothing once it runs.

Security tierTier 1, Contained.No network access and no credentials. Runs entirely on its own. What tiers mean

Install-time hooks & dependenciesno flags

Code that runs when you install it, before you ever call a tool.

Runs code / shell commandsno flags

Whether the server can execute commands on your machine.

Secrets & credentialsno flags

How it reads, logs, or transmits keys and tokens. Scam/wallet-drainer patterns land here.

Network calls outno flags

Hardcoded endpoints it reaches beyond what it documents.

Prompt-injection passthrough1 finding

INFOserver.py — HEURISTIC: this file both fetches external content and returns content as tool output, with no obvious sanitization. External text returned into tool output can carry instructions an agent obeys (prompt-injection passthrough). Confirm manually; this is a hint, not proof.

Permission scope breadthno flags

How much access it asks for versus what its job needs.

How to read this: these are static checks over the source at a point in time. They catch the patterns above, not everything. Absence of a flag is not absence of danger, and a tool that runs cleanly can still behave differently once installed. We do not call any tool simply "safe". Runtime-behavior checks are the next layer we are adding.

04RELATEDWORKS ALONGSIDE THIS

From the same session