Delete a Gmail message (move to trash)
AI agents call gmail_delete_message to permanently remove resources in Google — typically in cleanup and lifecycle workflows. It does its job in a single call, and there is no undo.
Deleting email messages is irreversible and represents data loss. Although technically moved to trash (which may have recovery options in Gmail), the primary action is destructive removal of a message from the user's active mailbox. This falls under the Destructive category as it eliminates data that cannot be undone through normal tool operations.
From the tool's definition The tool name is 'gmail_delete_message' and description states 'Delete a Gmail message (move to trash)' — this permanently removes user data from the inbox.
Attacks that exploit this kind of access
Delete a Gmail message (move to trash). It is categorised as a Destructive tool in the Google MCP Server, which means it can permanently delete or destroy data. Block by default and require explicit approval.
Register the Google MCP server in PolicyLayer and add a rule for gmail_delete_message: 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 Google. Nothing to install.
gmail_delete_message is a Destructive tool with critical risk. Critical-risk tools should be blocked by default and only enabled with explicit human approval.
Yes. Add a rate_limit block to the gmail_delete_message 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 gmail_delete_message. 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.
gmail_delete_message is provided by the Google MCP server (robcerda/google-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 →