Bring your own agent.

• A MySendbox team with Agents turned on, and at least one mailbox.
• An owner on the team to approve the connection.
• For the agent to reply, not only read: open the mailbox's settings in the dashboard and set Who answers to “Your connected agent” or “Either”.

No account yet? Add the server in your app anyway. When MySendbox asks you to sign in, choose “Create a workspace for your agent”: you sign up, pick an address for your agent (on agentsendbox.com, so no domain is needed), and approve the connection in one go. Your agent can read straight away; replies are sent once you start the 7-day trial ($1).

The server speaks Streamable HTTP and signs in with OAuth, so most apps only need the URL.

https://mcp.mysendbox.com/mcp

• Claude (web and desktop): Settings → Connectors → Add custom connector, and paste the URL.
• ChatGPT: add it as a custom connector (MCP server) with the same URL.

claude mcp add --transport http mysendbox https://mcp.mysendbox.com/mcp

{
  "mcpServers": {
    "mysendbox": { "url": "https://mcp.mysendbox.com/mcp" }
  }
}

The first time the agent connects, your app opens a page on app.mysendbox.com. Sign in as an owner, pick the mailboxes the agent may use, and choose its access:

• Read: read and search mail.
• Draft: read, tidy, and write drafts for a person on your team to send.
• Reply: read, tidy, and reply. Replies are checked, and held for a person when they're risky.

read    list_mailboxes  list_messages  search_messages
        get_message  get_thread  get_attachment  get_reply
        account_status   → can it send, credits, what to do if not
draft   + mark_read  move_messages  draft_reply
reply   + send_reply   → sent, or held for a person

There's no push: to hear about new mail, poll list_messages with since, or have a message.received webhook wake your agent.

Servers and scripts can skip OAuth and send an API key as the bearer token. Create one in the dashboard under Integrations, with the scopes the agent needs, or let the agent sign itself up and get a key with its own inbox:

Authorization: Bearer msb_live_…

• messages:read to read, drafts:write to draft, replies:send to reply.
• messages:write to tidy: mark read and move messages.

• Replies, not outreach. Agents answer people who wrote in, in the same thread, one reply per message. They can't add recipients or start new conversations.
• Same checks, same approvals. Every reply is screened for leaks and for instructions planted in the email. Risky ones wait for a person, whichever agent wrote them.
• Outside mail stays outside. Messages reach the agent marked as untrusted, with their screening results. Quarantined mail is never shown.
• Every call on the record. Each tool call is logged with the agent, mailbox and outcome. Connections are limited to 120 requests a minute, on top of each mailbox's daily limits.

More on how we keep email wanted: Safety.

Any owner can see every connected agent, its last use and its activity under Integrations → Connected agents in the dashboard, and disconnect it there. It loses access at once. Removing the person who connected it does the same.