AI agents call find_unread to retrieve information from Mail without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool queries existing email data and presents it without creating, modifying, deleting, executing code, or committing financial actions. It is a straightforward read operation on email metadata. The low severity reflects that misuse would only expose message visibility information without broader system impact.
From the tool's definition Tool is described as listing unread messages in a folder, with the equivalent operation being list_messages(unseen_only=True). The description uses the verb 'list' which retrieves data without modification.
Attacks that exploit this kind of access
Shortcut: list unread messages in a folder. Equivalent to list_messages(unseen_only=True). It is categorised as a Read tool in the Mail MCP Server, which means it retrieves data without modifying state.
Register the Mail MCP server in PolicyLayer and add a rule for find_unread: 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.
find_unread 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 find_unread 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 find_unread. 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.
find_unread 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 →