AI agents call search_messages 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.
Despite the empty description, the tool name and server context strongly indicate this is a query/search operation that retrieves email data without modifying it, placing it in the Read category. Severity is medium rather than low because email often contains sensitive personal and financial information; an agent could be tricked into searching for and exfiltrating private credentials or sensitive data.
From the tool's definition Tool name 'search_messages' combined with server description stating it 'enables AI agents to interact with email accounts for reading, searching, and sending messages' and listing 'search' as a core capability.
Attacks that exploit this kind of access
search_messages. 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 search_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.
search_messages 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 search_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 search_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.
search_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 →