Fetch the transcript/subtitles of a YouTube video with timestamps. First tries YouTube captions (instant, already timestamped). If no captions exist, downloads the audio and transcribes locally with Whisper, producing segment-level timestamps AND per-token confidence — the output includes Uncerta...
AI agents call get_youtube_transcript to retrieve information from Media without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
Even though get_youtube_transcript only reads data, uncontrolled read access leaks sensitive information and racks up API costs — an agent caught in a retry loop can make thousands of calls a minute without anyone noticing.
Attacks that exploit this kind of access
Fetch the transcript/subtitles of a YouTube video with timestamps. First tries YouTube captions (instant, already timestamped). If no captions exist, downloads the audio and transcribes locally with Whisper, producing segment-level timestamps AND per-token confidence — the output includes Uncertainty zones (spans where Whisper is guessing, with midpoint_s timestamps) and Demonstrative phrases that often reference on-screen content. When those appear and matter to the user. It is categorised as a Read tool in the Media MCP Server, which means it retrieves data without modifying state.
Register the Media MCP server in PolicyLayer and add a rule for get_youtube_transcript: allow, deny, rate-limit, or require approval. Point your MCP client at the PolicyLayer proxy URL and the rule is enforced on every call, before it reaches Media. Nothing to install.
get_youtube_transcript is a Read tool with low risk. Read-only tools are generally safe to allow by default.
Yes. Add a rate_limit block to the get_youtube_transcript rule in your PolicyLayer policy. For example, setting max: 10 and window: 60 limits the tool to 10 calls per minute. Rate limits are tracked per agent session and reset automatically.
Set action: deny in the PolicyLayer policy for get_youtube_transcript. The AI agent will receive a policy violation error and cannot call the tool. You can also include a reason field to explain why the tool is blocked.
get_youtube_transcript is provided by the Media MCP server (woosal1337/media-mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.