Sync InfraForge mailboxes to Smartlead every night

Build a code workflow that runs every night at 2am (workspace local time) and syncs newly provisioned InfraForge mailboxes into Smartlead as sending accounts. This is a deterministic pipeline. No LLM reasoning nodes.

Trigger: cron, daily at 02:00.

Configurable inputs on the workflow:

- infraforge_workspace_id: the InfraForge workspace to sync mailboxes from.

- smartlead_campaign_id: optional. When set, newly created Smartlead sending accounts are attached to this campaign.

- slack_channel: Slack channel or user id to post the summary to.

Steps:

1. Call infraforge: List Mailboxes scoped to the configured workspace_id, with with_credentials=true so SMTP and IMAP credentials come back in the response.

2. Call smartlead: List Email Accounts to fetch the current sending accounts in the Smartlead workspace.

3. Diff the two lists by email address (case-insensitive, trimmed). Build a list of InfraForge mailboxes whose email is not already present in Smartlead.

4. For each mailbox in the missing list, call smartlead: Create Email Account using the SMTP host, port, username, password, and IMAP host, port, username, password returned by InfraForge. Also carry over the from-name and any signature if provided. Collect the created email_account_ids and any per-mailbox failures.

5. If smartlead_campaign_id is set and at least one account was created successfully, call smartlead: Add Email Account to Campaign with the campaign id and the list of newly created email_account_ids.

6. Call slackbot: Send a Message to the configured Slack channel with a one-line summary: how many accounts were added, whether they were attached to a campaign, and any failures (with the offending email addresses and the error message).

Implementation notes: Use the slackbot integration (bot token) for the notification, not the user slack integration. Wrap each per-mailbox Create Email Account call so a single failure does not abort the whole run. Retry transient 429/5xx errors with a short backoff. Always post the Slack summary at the end, even when nothing was added, so operators know the job ran.