AgentPod is building a private, secure device for your AI agent.The AgentPod device is coming.Coming soonBe first
We checked this and rejected itsecurity

docx-mcp: why it is not safe to use

docx-mcp passed our review; we found local DOCX processing, no hidden data transfer, and no hardcoded secrets in inspected files.

What we found

Our review did not identify a security or privacy issue that would block use of `docx-mcp`. In the inspected README, install script, and server entrypoints, we did not find hidden instructions, disguised behavior, or intentionally obscured payloads. The repository materials indicate local DOCX processing and local skill installation, and we did not find code that sends document contents to an undisclosed third party. We also did not find hardcoded credentials, API keys, or token material in the inspected files.

We noted that the skill can overwrite files and perform permanent redaction when explicitly invoked. Based on the inspected code, we did not find evidence of automatic irreversible actions without a user-directed tool call. We also did not find runtime remote code fetch and execution in the Python entrypoints, although the README does disclose an optional `curl`-to-shell installer alongside standard package-manager installation.

What to do instead

If you use this skill, prefer the standard package-manager install path over any `curl`-to-shell option, review file paths before running write or redaction actions, and test on copies of documents when possible.

Want the same outcome, safely? Use our checked skill instead.

Source: https://github.com/SecurityRonin/docx-mcp

We report what our security review found at the time we checked, with the goal of keeping people safe. Projects change; if a maintainer has since fixed this, we are glad to recheck it. Email hello@agentpod.com.

Copied to clipboard. Paste it into your AI (ChatGPT, Claude, or your agent) to add the skill.