Purpose: Send a single transactional email to one or more recipients immediately (or schedule it). Use for one-off messages, notifications, and direct replies. NOT for: Sending the same email to a whole list/audience (use create-broadcast + send-broadcast). Not for managing contacts or audiences....
AI agents use send-email to create or update resources in Email Sending MCP — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Email Sending MCP environment.
This tool creates new email messages and dispatches them, which constitutes a Write operation (data creation with side effects). It is not Read (it modifies state by sending emails), not Execute (it doesn't run arbitrary code or shell commands), not Destructive (emails can be unsent or retracted in some systems, and the primary harm is communication-based rather than data loss), and not Financial (no money movement).
From the tool's definition Tool description states it "Send[s] a single transactional email to one or more recipients immediately (or schedule it)" and "Returns: Send confirmation and email ID." The action creates/sends data (email messages) with reversible side effects.
Documented attack patterns abuse exactly the kind of access send-email gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and Email Sending MCP, and nothing reaches the server without passing your rules. This is the rule we recommend for send-email:
{
"version": "1",
"default": "deny",
"tools": {
"send-email": {
"limits": [
{
"counter": "send-email_rate",
"window": "minute",
"max": 30,
"scope": "grant"
}
]
}
}
}send-email stays usable, but capped — an agent stuck in a loop can't make hundreds of changes a minute. Everything else on the server is denied unless you say otherwise.
Free to start. No card required.
Purpose: Send a single transactional email to one or more recipients immediately (or schedule it). Use for one-off messages, notifications, and direct replies. NOT for: Sending the same email to a whole list/audience (use create-broadcast + send-broadcast). Not for managing contacts or audiences. Returns: Send confirmation and email ID. When to use: - User wants to. It is categorised as a Write tool in the Email Sending MCP MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the Email Sending MCP server in PolicyLayer and add a rule for send-email: 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 Email Sending MCP. Nothing to install.
send-email 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 send-email 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 send-email. 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.
send-email is provided by the Email Sending MCP server (resend/resend-mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Deterministic rules across all 77 Email Sending MCP tools. Per-identity grants. Full audit log. Live in minutes. Nothing to install.
Free to start. No card required.
77 Email Sending MCP tools catalogued and risk-classified — across an index of 42,500+ MCP servers.