AI agents use save_draft to create or update resources in Mail — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Mail environment.
The tool creates or modifies draft messages reversibly without sending them. While the description is empty (lowering confidence slightly), the name and server purpose (email management with SMTP support) clearly indicate this creates/stores email data. This is Write rather than Execute because it does not trigger external operations—drafts remain local to the email system until explicitly sent.
From the tool's definition Tool name 'save_draft' combined with server context indicating SMTP/email management capabilities; drafts are created/modified email messages that persist in the email system.
Attacks that exploit this kind of access
save_draft. It is categorised as a Write tool in the Mail MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the Mail MCP server in PolicyLayer and add a rule for save_draft: 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.
save_draft 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 save_draft 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 save_draft. 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.
save_draft 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 →