Skip to content
The AI Agent Factory

Chapter 9 on today's products

Approved by
Panaversity ·
13 min read

This page backs 9.8, on running the same recurring work on today's tools. The lesson teaches the differences that change your setup: who the work runs as, what can start it and how its initiative is limited. This page holds each product's own settings, with their sources and the dates they were checked.

Products change every few weeks. Check the date on each box before you rely on a line, and open its source if the date is old.

The two AI vendors' lists use the same seven lines, in the same order. Each adds one of its own: Anthropic's shared worker in Slack, and OpenAI's specialist dot.

Anthropic, as verified 18 November 2026.

  • Schedules. Scheduled tasks in Claude run at set times, on paid plans.1 They run remotely. As of 18 November 2026, a task with a folder on your own computer has "Require this computer" turned on. It runs only while that computer is awake and Claude Desktop is open.1 A cloud session reaches your folders only while the desktop app is open.2 So check whether your task requires this computer. Chat and Cowork are one Claude, which is rolling out in stages on the Pro, Max and Ultra plans. On Team and Enterprise, it is available where the Chat and Cowork unified setting is on.3 Claude Code, the coding harness, has scheduled tasks of its own, in research preview, that add custom schedules. Until recently it called them routines.4
  • Event triggers. Claude Code's scheduled tasks can start from an API call or from GitHub events.4 Anthropic's pages do not describe a task in the Claude app that starts when an email arrives.
  • Each run. Each scheduled task runs as its own session.1 If the trigger of a Claude Code task is sent again, it starts another session.5 A run marked Succeeded ended without Claude Code reporting a failure. It does not mean Claude did the work you asked for.4
  • Whose name. A scheduled task in Claude Code belongs to your own account and acts as you.4 In a channel, Claude Tag, a shared Claude in a Slack channel, works under its own identity, with the access an admin sets for that channel. In a one-to-one direct message, it works on that member's own Claude account. In a group message, it keeps its own identity. In a channel, it uses a member's own connectors only for that member's task, once the member allows it.6 Claude Duet runs on your Claude account, and with your approval it sends email from the accounts you connect.7
  • Approvals. A scheduled task has an approval mode.1 The modes run from manual approval to "Skip all approvals." Even with "Skip all approvals," permanent file deletion always needs your permission.8 A run of a Claude Code scheduled task starts in auto mode, where automated checks allow or block each action. It can still stop and wait for your answer. So you control these tasks mainly by what they can reach.4
  • Standing worker, verified 18 November 2026. Claude Duet is Anthropic's standing worker, in research preview, with a waitlist for Pro, Max and Ultra subscribers in the US. It checks your connected accounts every few hours and sends a short brief each morning. It messages you only when something needs your attention. Before it sends a message, makes a payment, signs a document or deletes anything, it waits for your approval.9
  • Shared worker in Slack. Claude Tag, in beta on Team and Enterprise, is one shared Claude in a Slack channel.10 It follows up on its own.10 It can reply without being tagged, but only while "Respond automatically" is on.11
  • Review and stop. Past runs are on the Scheduled page, where a task can be paused.1 Claude Tag's Activity page lists every task for Owners, and its spending can be limited.10

OpenAI, as verified 21 November 2026.

  • Schedules. Scheduled tasks run once or on a schedule.12 How often depends on the kind of task: a task scheduled inside a chat can repeat at intervals measured in minutes.13 Your plan limits how many tasks can be active.12 Team tasks, on Business and Enterprise, are recurring work owned by a team.14
  • Event triggers. Tasks in ChatGPT can start from Gmail, Slack or GitHub activity. On Enterprise and Edu, an admin must allow this first.12
  • Each run. A scheduled task that is not inside a chat starts a new chat each time it runs. A task scheduled inside a chat returns to that chat.13 A team task keeps a list of previous runs.14
  • Whose name. Team tasks run under the team's service account, an account that belongs to the team, not to the creator.14
  • Approvals. Approval rules in connected apps still apply, and an action that needs approval may pause a task.12
  • Standing worker. A dot is an always-on agent with its own cloud computer and memory.15 Dots come with Pro and Business Premium, and are reaching some Plus users. Enterprise and Edu workspaces have them unless an admin opts out.16 Its own background research uses read-only tools, which can look but not change anything.15 For some actions, its Custom Rules let you choose "take action without asking," "take action when you say so," "ask before taking action" or "hand off to you."17
  • Review and stop. Activity View shows a dot's work.15 Pause stops its current main task, but not every delegated task, and not its scheduled runs. Delete does not undo changes already made in connected apps.17
  • A specialist dot, verified 21 November 2026. A company can set one up for dedicated work, with its own identity, credentials and access to the systems it needs. OpenAI's own early testing included invoice processing. It is starting with focused enterprise pilots, in which its engineers help define the dot's responsibilities and how people review and approve its work.15

Open source, as verified 23 October 2026.

  • Schedules. OpenClaw, an open-source assistant that runs on your own computer,18 has a scheduler called Automations. It runs a job once or on a cron schedule, a timetable that repeats.19 Omnara, a runtime that you can host yourself,20 has cron triggers that run in the time zone you choose.21
  • Event triggers. OpenClaw's Automations also cover webhooks, messages that another system sends when something happens, and Gmail triggers.19
  • Each run. Each time it fires, an Omnara trigger can start a new agent from a saved profile. Or it can send input to an agent that is already running.21
  • Standing worker. OpenClaw is a standing assistant.18 Its first setup creates a "Proactive check" job, which runs every 30 minutes by default, in the agent's main conversation.22
  • Running it safely. Run OpenClaw on a dedicated machine, virtual machine or container,23 and give it only the accounts and permissions it needs. Its docs say to avoid opening it directly to the internet.24

Check yourself

Question 1 / 1

0 answered

Which four settings do a dot's Custom Rules let you choose?

  1. Schedule recurring tasks in Claude Cowork, Claude Help Center.↑↑2↑3↑4↑5
  2. Use Claude Cowork on web, desktop, and mobile, Claude Help Center.↑
  3. Claude Cowork and chat are one Claude, Claude Help Center.↑
Owner
Panaversity

On this page