AI agents call youtube_list_captions to retrieve information from Youtube without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool only retrieves and enumerates caption metadata for a video. It performs a query operation with no side effects, no data modification, and no external code execution. The blast radius of misuse is minimal—an agent could only enumerate captions, which is publicly discoverable information about the video. This is a straightforward Read category tool.
From the tool's definition Tool name contains 'list' and description states 'List all caption tracks available for a YouTube video' - a pure data retrieval operation with no modification, deletion, or execution capability.
Attacks that exploit this kind of access
List all caption tracks available for a YouTube video. Args - \. It is categorised as a Read tool in the Youtube MCP Server, which means it retrieves data without modifying state.
Register the Youtube MCP server in PolicyLayer and add a rule for youtube_list_captions: 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 Youtube. Nothing to install.
youtube_list_captions 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 youtube_list_captions 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 youtube_list_captions. 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.
youtube_list_captions is provided by the Youtube MCP server (tuitamogamer-gpt/youtube-mcp-server). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Every MCP server has a record like this.
Type a name, get the same breakdown: verified identity, auth posture, risk grade, capabilities, recommended policy.
Teams ship this data inside their own products. See what a licence covers →