YouTube Transcript MCP: why it is not safe to use
We did not find security or privacy issues in this skill during review, and it passed with limited, transcript-related network access.
What we found
Our review of **YouTube Transcript MCP** did not identify a security or privacy issue that would block use in its stated role. In the inspected repository, we found no hidden instructions, hardcoded secrets, obfuscated code, or commands that delete, overwrite, or send data irreversibly. Observed network behavior was limited to fetching YouTube watch, caption, InnerTube, and oEmbed data needed to retrieve transcripts and related metadata. The repository also states these requests are made from the user's own machine.
We also did not find code that accesses browser data, keychains, SSH keys, or unrelated environment variables. The code checks only `NODE_ENV` for debug logging. We found no path for remote code execution or privilege escalation, and the Dockerfile switches to a non-root user at runtime.
What to do instead
This skill **passed** our review with a score of **92/100**. If you use it, keep it scoped to transcript retrieval, review any future updates, and continue standard controls such as least-privilege runtime settings, dependency monitoring, and routine network review for any changes in behavior.
Source: https://github.com/sinco-lab/mcp-youtube-transcript
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.