Classify inbox mail into keep, flag, archive, and review buckets.
AI agents call mail_audit to retrieve information from Personal Mail without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
The tool reads and classifies existing inbox mail into categories (keep, flag, archive, review). This is primarily a read/analysis operation — it retrieves messages and assigns them to buckets. However, if classification triggers downstream actions (e.g., auto-archiving), it could be Write; the description does not confirm side effects, so Read is the best fit.
From the tool's definition Classify inbox mail into keep, flag, archive, and review buckets
Attacks that exploit this kind of access
Classify inbox mail into keep, flag, archive, and review buckets. It is categorised as a Read tool in the Personal Mail MCP Server, which means it retrieves data without modifying state.
Register the Personal Mail MCP server in PolicyLayer and add a rule for mail_audit: 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 Personal Mail. Nothing to install.
mail_audit 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 mail_audit 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 mail_audit. 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.
mail_audit is provided by the Personal Mail MCP server (srogerf/personal-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 →