// SCAN REPORT

Mailbox

49 tools. 24 can modify or destroy data without limits.

5 destructive tools with no built-in limits. Policy required.

Last updated: 6 May 2026

24 can modify or destroy data

25 read-only

49 tools total

// TOOL MAP

Read (25) Write / Execute (19) Destructive / Financial (5)

// WHAT CAN GO WRONG

Destructive tools (delete_draft, delete_filter, delete_label) permanently delete resources. There is no undo. An agent calling these in a retry loop causes irreversible damage.

Write operations (archive_email, authenticate, batch_modify_emails) modify state. Without rate limits, an agent can make hundreds of changes in seconds — faster than any human can review or revert.

Execute tools (reauth) trigger processes with side effects. Builds, notifications, workflows — all fired without throttling.

// RECOMMENDED RULES

Deny destructive operations

delete_draft:
  rules:
    - action: deny

Destructive tools should never be available to autonomous agents without human approval.

Rate limit write operations

archive_email:
  rules:
    - rate_limit: 30/hour

Prevents bulk unintended modifications from agents caught in loops.

Cap read operations

bulk_trash:
  rules:
    - rate_limit: 60/minute

Controls API costs and prevents retry loops from exhausting upstream rate limits.

// ALL 49 MAILBOX TOOLS

DESTRUCTIVE 5 tools

Destructive delete_draft Destructive delete_filter Destructive delete_label Destructive delete_template Destructive remove_account

EXECUTE 1 tools

Execute reauth

WRITE 18 tools

Write archive_email Write authenticate Write batch_modify_emails Write bulk_modify Write create_draft Write create_filter Write create_label Write export_email Write export_thread Write mark_read Write modify_email Write save_template Write send_draft Write send_email Write send_template Write set_signature Write set_vacation Write update_draft

// FAQ

Can an AI agent delete data through the Mailbox MCP server? +

Yes. The Mailbox server exposes 5 destructive tools including delete_draft, delete_filter, delete_label. These permanently remove resources with no undo. Intercept blocks destructive tools by default so they never reach the upstream server.

How do I prevent bulk modifications through Mailbox? +

The Mailbox server has 18 write tools including archive_email, authenticate, batch_modify_emails. Set rate limits in your policy file -- for example, rate_limit: 10/hour prevents an agent from making more than 10 modifications per hour. Intercept enforces this at the transport layer.

How many tools does the Mailbox MCP server expose? +

49 tools across 4 categories: Destructive, Execute, Read, Write. 25 are read-only. 24 can modify, create, or delete data.

How do I add Intercept to my Mailbox setup? +

One line change. Instead of running the Mailbox server directly, prefix it with Intercept: intercept -c mailbox.yaml -- npx -y @mailbox-mcp. Download a pre-built policy from policylayer.com/policies/mailbox and adjust the limits to match your use case.

// RELATED SERVERS

Other MCP servers with similar tools.

Starter policies available for each. Same risk classification, same one-command setup.

Mailbox

Other MCP servers with similar tools.

Let agents act without letting them run wild.