OpenClaw unsubscribes forwarded vendor emails

After re:Invent, Vivek V. received vendor follow-up campaigns that were tedious to unsubscribe from individually. He gave an OpenClaw agent its own email inbox and forwarded vendor messages there through a mail rule. The agent parsed each message, found the unsubscribe link, followed it and confirmed the action.

Vivek reports clearing 47 vendor lists in two weeks. The separation of inboxes is part of the actual setup: he explicitly says the agent did not receive access to his personal mailbox. It operated on messages forwarded into its own account.

The account places this agent within a fleet of eighteen. Vivek built an AgentOps control system using Notion for an agent registry, task queue, run log and alerts. Operators could pause dispatch by editing an agent's status, inspect failures and synchronize configurations. Those controls describe how he managed the fleet; the detailed fleet counters are not presented as measurements of the unsubscribe agent alone.

This is an author's reported deployment, not an independent test of every destination or a universal way to stop spam. The article does not document which lists required login, how confirmation pages varied or whether every removal remained effective later. Its concrete outcome is a repeated unsubscribe workflow on a bounded inbox, with a reported two-week count and an operational way to inspect and pause agents.

Prompt
Suggested prompt (written by Agents In Use, not the author's)

Source: DEV Community · Vivek V.