Short tutorials for
common agent tasks.

How to forward email

You want the agent to process mail that was sent to your real inbox, without handing it OAuth access. Forwarding rules are the answer.

Read →

How to give an AI agent

Not an alias, not a shared mailbox, not your Gmail. A real inbox the agent owns and reads from on its own identity — set up in about a minute.

Read →

How to handle 2FA

When the login screen asks for a six-digit code, the agent should answer — without waking you up at 3am to read your authenticator app.

Read →

How to let an AI agent

Sign-up flows assume a human with an inbox and a phone. Here's how to give the agent both — and let it complete the flow without a person in the loop.

Read →

How to make an AI agent

Correct threading is the difference between an agent that looks like an assistant and an agent that looks like spam. Here's how to get it right.

Read →

How to make Claude

Tell Claude to send an email and it does — by pointing Claude Desktop or the Claude Agent SDK at a Loomal MCP server.

Read →

How to parse an email reply

Raw email bodies are a mess of quoted history, signatures, and disclaimers. Here's how to hand the model only the content it needs.

Read →

How to receive

Let your agent act on incoming mail the moment it arrives — no polling, no IMAP, just an HTTP endpoint Loomal calls when mail lands.

Read →

How to send email

The shortest working recipe for making a Node.js agent send email from its own DKIM-signed address. No SMTP, no OAuth.

Read →

How to send email

The shortest working recipe for making a Python agent send email from its own address — no SMTP setup, no OAuth, just an API key.

Read →

How to store API keys

Not in .env. Not in a JSON config. Not in a shared secrets manager with wildcard scopes. Here's what actually works for autonomous agents.

Read →

How to verify who

An agent that acts on email content is vulnerable to spoofed senders. Here's how to check DKIM, SPF, and DMARC before trusting a message.

Read →