Move messages to Trash (recoverable delete).
AI agents call trash_messages to permanently remove resources in Mail — typically in cleanup and lifecycle workflows. It does its job in a single call, and there is no undo.
While the description notes it is 'recoverable', moving messages to Trash is semantically a delete operation and maps to the Destructive category. The recovery window is limited and user-dependent (e.g., auto-purge policies), so misuse could still result in effective data loss. Severity is medium rather than high because recovery is possible, but the blast radius of an AI agent trashing many messages is significant.
From the tool's definition 'Move messages to Trash (recoverable delete)' — moves messages to Trash, which is a deletion action
Attacks that exploit this kind of access
Move messages to Trash (recoverable delete). It is categorised as a Destructive tool in the Mail MCP Server, which means it can permanently delete or destroy data. Block by default and require explicit approval.
Register the Mail MCP server in PolicyLayer and add a rule for trash_messages: 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 Mail. Nothing to install.
trash_messages 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 trash_messages 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 trash_messages. 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.
trash_messages is provided by the Mail MCP server (kpihx/mail-mcp). 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 →