Skip to main content
Agent self-signup allows an autonomous AI agent to provision its own persistent identity, dedicated email address ({agent}@wireboxmail.com), and API key in a single HTTP request without requiring a human to manually create an account first. The agent provides the email of the human developer who oversees it. The human receives a 6-digit verification code. Once verified, full sending capabilities are unlocked, and the human can manage the agent directly from the Wirebox Console.

Copy Prompt for your Agent

Copy and paste this prompt directly into your AI agent’s chat, terminal, or IDE (Claude Code, Cursor, OpenAI Codex, OpenClaw, Hermes, etc.):

Capabilities & Sandbox

To prevent spam and abuse while giving the agent immediate utility, new signups start in a protected sandbox until verified:

Before Verification (Unclaimed)

  • Allowed send recipients: Human developer’s email only (up to 3 test messages).
  • Outbound limit: 3 messages per 24 hours.
  • Inbound mail receiving: Fully enabled ({agent}@wireboxmail.com).
  • Create extra identities/mailboxes: Disabled.

After Verification (Claimed)

  • Allowed send recipients: Anyone on the internet.
  • Outbound limit: Plan-based (Free tier: 100/month).
  • Inbound mail receiving: Fully enabled.
  • Create extra identities/mailboxes: 1 dedicated mailbox per identity.
The human developer receives only one single official verification email from Wirebox (Wirebox <noreply@wireboxmail.com>). The agent communicates directly with the developer in its active terminal or IDE chat session to request the code.

End-to-End Walkthrough

Follow these sequential steps to register and verify your agent.

1. Register with Wirebox

Call the public registration endpoint. No authorization header or existing API key is needed.
Response (201 Created):
Store your API key securely: The api_key (wb_live_...) is returned only once upon registration. Set it in your runtime environment (e.g. WIREBOX_API_KEY).
Identity-Scoped API Key: For enterprise security, the API key issued during self-signup is strictly identity-scoped (messages:read, messages:send, mailboxes:read). It cannot create new API keys, access other agent mailboxes, or alter organization settings. Global admin keys can only be minted in the Developer Console via human authentication.

2. Verify with the 6-Digit Code

Upon registration, Wirebox dispatches an official verification email from Wirebox <noreply@wireboxmail.com> directly to the specified human_email. The agent prompts the human developer directly in the chat or terminal:
“I’ve provisioned my Wirebox mailbox (agent-7k9x2m@wireboxmail.com). A 6-digit verification code was sent to developer@example.com. Please enter the code here to complete setup.”
Once the developer enters the 6-digit code, submit it to claim the identity and unlock full capabilities:
Response (200 OK):

3. Send & Receive Messages (Full Capabilities)

With verification complete, sandbox restrictions are removed. The agent can now send outbound emails to any address and check incoming messages:

Sending Outbound Emails

Reading Inbox Messages


Troubleshooting & Lifecycle

Wirebox enforces a 60-second cooldown between verification email dispatches. If you need a new code, wait 60 seconds and call POST /api/v1/agent-signup again with the same human_email. Wirebox will refresh the OTP code, rotate the API key, and resend the email automatically.
Codes expire after 15 minutes. Calling POST /api/v1/agent-signup with the same human_email will generate a fresh code.
When the human developer subsequently logs into the Wirebox Web Console using Google or GitHub OAuth with the same email, the agent’s organization and mailbox are automatically linked with Owner permissions.

Next Steps

  • Email — Learn how to send HTML emails, manage multi-turn threads, and handle attachments.
  • Webhooks — Configure webhooks to receive instant HTTP notifications when new emails arrive.