AI agents use youtube_update_caption to create or update resources in Youtube — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Youtube environment.
This tool modifies existing captions reversibly—the changes can be undone by updating again or restoring from backups. It does not delete data (which would be Destructive) nor execute arbitrary code (Execute). While it affects a public YouTube channel's content, the blast radius is limited to caption metadata for a single video, and the action is reversible.
From the tool's definition Tool name contains 'update' and description states 'Update an existing caption track', indicating modification of existing data. The caption track is metadata associated with a video.
Attacks that exploit this kind of access
Update an existing caption track. It is categorised as a Write tool in the Youtube MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the Youtube MCP server in PolicyLayer and add a rule for youtube_update_caption: 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_update_caption is a Write tool with medium risk. Write tools should be rate-limited to prevent accidental bulk modifications.
Yes. Add a rate_limit block to the youtube_update_caption 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_update_caption. 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_update_caption 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 →