Creates a new email draft without sending it
AI agents use create-draft to create or update resources in Outlook Assistant — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Outlook Assistant environment.
This tool creates new email draft objects in the user's mailbox, which is a reversible write operation. While drafts can be deleted, the creation itself is not destructive. Severity is medium because a compromised agent could create many drafts, potentially for phishing or spam purposes, but the impact is limited since drafts are not automatically sent and remain in the user's control for review/deletion.
From the tool's definition Tool description states it 'Creates a new email draft without sending it' — a create operation that modifies data (drafts mailbox) reversibly.
Attacks that exploit this kind of access
Creates a new email draft without sending it. It is categorised as a Write tool in the Outlook Assistant MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the Outlook Assistant MCP server in PolicyLayer and add a rule for create-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 Outlook Assistant. Nothing to install.
create-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 create-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 create-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.
create-draft is provided by the Outlook Assistant MCP server (titanzero/outlook-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 →