superagnt_
// claude + superagnt_

Can Claude run tasks on a schedule?

Yes. Claude's scheduled tasks run recurring prompts on paid plans, in Anthropic's cloud even while your computer sleeps. superagnt_ adds schedules for hosted agents that any MCP client can set up, Claude, ChatGPT or Codex, on any cron and time zone, with your superagnt_ data.

In the run below, Claude built a small agent and gave it a weekday 9:00 schedule in New York time. At 9:00 it fired with no client connected, read a table in the workspace database and answered in 11 seconds.

Connect superagnt_ to Claude

Copy the setup prompt.
Your agent does the rest.

works with any agent+ 8 more
Install superagnt for me. It's the superagnt plugin from the github.com/superagnt/plugins marketplace, plus the superagnt MCP server (on Claude it's a connector I add in my browser; elsewhere I sign in through my browser). The setup guide for each kind of agent is at https://www.superagnt.com/install
or follow the detailed setup instructions

» the base connection already serves agents, schedules and their run history. Nothing to switch on.

01// on its own, and with superagnt_

What Claude does alone, and what changes once it's connected

jobClaude on its ownwith superagnt_
Recurring tasksYes: scheduled tasks in Cowork, on every paid plan, hourly, daily, on weekdays or weekly. They run remotely, so a sleeping computer doesn't stop them, unless a task needs local files.A hosted agent on any five-field cron, read in the IANA time zone you name per schedule.
From Claude CodeCloud scheduled tasks (research preview, Pro and up) run on Anthropic's infrastructure. Cron is read in UTC unless you set CRON_TZ, and nothing runs more than once an hour.The schedule is a tool on the MCP server (agnt_schedules_write), so Claude Code, ChatGPT, Codex and Cursor all set it up the same way.
What a run can useYour connectors, skills, plugins and the files in your Claude account.superagnt_'s data sources and your workspace database: the run below read a table at 9:00.
02// a real run

The job, run on superagnt_

Real tool calls on superagnt_'s staging environment, the same calls Claude makes once the server is connected. Each result is what the client received, trimmed where it says so and otherwise unchanged, and each wall time was measured around the call.

the ask

“Build an agent that reads my signups table every weekday at 9am New York time and reports the count and the newest signup.”

  1. 01Create the agent356 ms wall time
    → callagnt_agents_write
    {
      "action": "create",
      "name": "Weekday signup digest",
      "description": "Reads the signups table every weekday morning and reports what came in.",
      "system_prompt": "You run on a schedule. Each run, read the signups table in the workspace database with agnt_db_read and reply in two lines: how many signups the table holds, and the newest one (name, company, date). No other text."
    }
    ← result
    {
      "id": "61291d32-4664-4dce-97dd-98867ebe6c00",
      "name": "Weekday signup digest",
      "description": "Reads the signups table every weekday morning and reports what came in.",
      "draft": true,
      "disabled": false,
      "archived": false,
      "model": "claude-sonnet-5",
      "default_channel": null,
      "tool_count": 0,
      "updated_at": "2026-10-26T12:43:52.353694+00:00",
      "next_step": "Draft saved. It does not run until `agnt_agents_write(action: \"deploy\")` is called. Configure it first with `agnt_agents_write(action: \"update_config\")`."
    }
  2. 02Let it read the workspace database617 ms wall time
    → callagnt_agents_write
    {
      "action": "update_config",
      "agent_id": "61291d32-4664-4dce-97dd-98867ebe6c00",
      "patch": {
        "workspace_db": {
          "enabled": true
        }
      }
    }
    ← result
    {
      "id": "61291d32-4664-4dce-97dd-98867ebe6c00",
      "name": "Weekday signup digest",
      "description": "Reads the signups table every weekday morning and reports what came in.",
      "draft": true,
      "disabled": false,
      "archived": false,
      "model": "claude-sonnet-5",
      "default_channel": null,
      "tool_count": 0,
      "updated_at": "2026-10-26T12:44:02.276+00:00",
      "applied": true,
      "patch_summary": [
        "workspace_db=merge"
      ],
      "note": "Applied to the draft. Deploy it to make the change live."
    }
  3. 03Deploy it3.0 s wall time
    → callagnt_agents_write
    {
      "action": "deploy",
      "agent_id": "61291d32-4664-4dce-97dd-98867ebe6c00"
    }
    ← result
    {
      "id": "61291d32-4664-4dce-97dd-98867ebe6c00",
      "name": "Weekday signup digest",
      "description": "Reads the signups table every weekday morning and reports what came in.",
      "draft": false,
      "disabled": false,
      "archived": false,
      "model": "claude-sonnet-5",
      "default_channel": null,
      "tool_count": 0,
      "updated_at": "2026-10-26T12:44:08.978+00:00",
    …
      "already_deployed": false,
      "note": "Deployed and live. It now consumes workspace credits when it runs."
    }

    » Trimmed: the agent's gateway address on staging.

  4. 04Schedule it for weekdays at 9:00 New York time967 ms wall time
    → callagnt_schedules_write
    {
      "action": "create",
      "agent_id": "61291d32-4664-4dce-97dd-98867ebe6c00",
      "name": "Weekday 9am signup digest",
      "cron_expression": "0 9 * * 1-5",
      "timezone": "America/New_York",
      "input_text": "Good morning. Read the signups table and report how many signups it holds and the newest one."
    }
    ← result
    {
      "schedule": {
        "schedule_id": "202553e0-3726-4212-bd1f-dec4884f9a2d",
        "deployed_agent_id": "61291d32-4664-4dce-97dd-98867ebe6c00",
        "name": "Weekday 9am signup digest",
        "cron_expression": "0 9 * * 1-5",
        "timezone": "America/New_York",
        "input_text": "Good morning. Read the signups table and report how many signups it holds and the newest one.",
        "enabled": true,
        "next_run_at": "2026-10-26T13:00:00+00:00",
        "last_run_at": null
      },
      …
    }

    » Trimmed: the closing note, which repeats the next run time and says each firing is a billed agent run.

  5. 05After 9:00: Check that it fired370 ms wall time
    → callagnt_observability_read
    {
      "action": "list_trigger_runs",
      "agent_id": "61291d32-4664-4dce-97dd-98867ebe6c00",
      "limit": 5
    }
    ← result
    {
      "agent_id": "61291d32-4664-4dce-97dd-98867ebe6c00",
      "trigger_runs": [
        {
          "trigger_type": "schedule",
          "schedule_id": "202553e0-3726-4212-bd1f-dec4884f9a2d",
          "agent_session_id": "95dc6dcc-a9af-4ab3-ac7e-8aa2b3ffd2b7",
          "status": "dispatched",
          "error": null,
          "created_at": "2026-10-26T13:00:04.876767+00:00"
        }
      ]
    }

    » Trimmed: internal ids on the run record.

  6. 06After 9:00: Read what the run said686 ms wall time
    → callagnt_observability_read
    {
      "action": "read_session",
      "session_id": "95dc6dcc-a9af-4ab3-ac7e-8aa2b3ffd2b7",
      "event_ids": [
        "sevt_01TZc7Fs2iWGY2iy1kkJkzQA",
        "sevt_01DNwPYZxYuvi1toAnzHPRnq",
        "sevt_018TAR7gSJi5zgUfjSkyuLq8",
        "sevt_01NFc15K1Pmo99KvY3Vpdxp9",
        "sevt_01PwW3e9StjPFJbPwAw1rhNs",
        "sevt_01AtfsHDEQqMxC5GA8YwQ6cU",
        "sevt_015R3bTiGPrdfQkKAjfmZKa6"
      ]
    }
    ← result
    {
    …
        {
          "index_position": 19,
          "type": "agent.message",
          "processed_at": "2026-10-26T13:00:17.305255Z",
          "text": "The signups table holds 7 records.\nThe newest signup is Gus Hale (COO, Proseware) on 2026-10-25.",
    …

    » Trimmed: the run's other 23 events, its four tool calls among them.

The schedule fired at 13:00:04 UTC, 9:00 in New York, with nothing of ours connected. The run took 11 seconds and four tool calls (the first guessed a column that doesn't exist; the agent looked again and carried on), and agnt_observability_read shows every firing, each tool call and the reply. A scheduled run is a background session, so its reply goes nowhere on its own: to hear about it, add notify_human to the agent's tools and have it send the result (it lands in your Slack, iMessage or superagnt_ app conversation with the agent), or have it write the result to a table.

03// set it up

Set it up in your client

Claude

  1. 01Open the prefilled Add custom connector dialog on claude.ai. One connector covers Claude on the web, Claude Desktop and Cowork.
  2. 02Click Add, then approve on the superagnt_ consent screen. It creates your superagnt_ account if you don't have one.
  3. 03Ask: “Build an agent that reads my signups table every weekday at 9am New York time and reports the count and the newest signup.”

Free plans can add one custom connector. On Team and Enterprise plans, an owner adds it under Organization settings, Connectors (Anthropic).

Claude Code

  1. 01In Claude Code, run /plugin marketplace add superagnt/plugins, then /plugin install superagnt@superagnt for the superagnt_ skills.
  2. 02Add superagnt_ to Claude as a connector with the prefilled dialog and click Add. Claude Code signed in with your Claude account picks it up.
  3. 03Start a new session.
  4. 04Ask: “Build an agent that reads my signups table every weekday at 9am New York time and reports the count and the newest signup.”

ChatGPT

  1. 01In ChatGPT on the web, go to Plugins, select the plus button, then Add custom MCP server.
  2. 02Name it superagnt, enter the server URL under Connection, and set Authentication to OAuth. The server URL is https://mcp.superagnt.com/mcp.
  3. 03Tick "I understand and want to continue", then select Create as a plugin. Sign in to superagnt_ when asked. No token needed.
  4. 04Install the plugin, then start a new chat, type @ and select superagnt.
  5. 05Ask: “Build an agent that reads my signups table every weekday at 9am New York time and reports the count and the newest signup.”

ChatGPT takes custom MCP servers on the web, not on mobile, and only Business and Enterprise/Edu workspaces get write actions. On Pro, connected servers are read and fetch only, so creating an agent and its schedule needs one of those plans in ChatGPT (OpenAI).

Codex

  1. 01In a terminal, run codex mcp add superagnt --url https://mcp.superagnt.com/mcp, then codex mcp login superagnt. The login opens a browser to approve.
  2. 02Start a new Codex session.
  3. 03Ask: “Build an agent that reads my signups table every weekday at 9am New York time and reports the count and the newest signup.”
04// side by side

Claude's scheduled tasks and superagnt_ schedules

Claude Cowork scheduled tasksClaude Code cloud scheduled taskssuperagnt_ agent schedules
PlansEvery paid planPro, Max, Team and Enterprise, in research previewEvery plan. Enabled schedules: Starter 10, Growth 50 and Scale 250
Runs while your computer sleepsYes, except tasks that need local files or appsYesYes
CadenceHourly, daily, weekdays, weekly or manualPresets or cron, at most once an hourAny five-field cron
Time zoneNot documentedPresets in your local time; cron in UTC unless CRON_TZ is setAny IANA time zone, set per schedule
Set up fromThe Claude appsClaude Code, the web or the desktop appAny MCP client: Claude, Claude Code, ChatGPT, Codex, Cursor
Claude's columns are from Anthropic's help center and Claude Code docs, checked Oct 26, 2026.

They work together, too: Anthropic's docs say cloud scheduled tasks use your connectors, so once superagnt_ is added to Claude as a connector, a Claude scheduled task can call its tools.

05// questions

Common questions

01Can Claude run tasks on a schedule?

Yes. Claude's scheduled tasks run recurring prompts on paid plans, in Anthropic's cloud even while your computer sleeps. superagnt_ adds schedules for hosted agents that any MCP client can set up, on any cron and time zone.

02Do Claude's scheduled tasks run when my computer is off?

Cowork's do: they run remotely, unless a task needs local files or apps. Claude Code's cloud tasks run on Anthropic's infrastructure. Claude Code's local desktop tasks run only while the app is open and the computer is awake.

03How often can a superagnt_ schedule run?

On any standard five-field cron expression, read in the time zone you set. Each firing is a billed agent run.

04How many schedules can I have on superagnt_?

Enabled schedules per plan: Starter 10, Growth 50 and Scale 250.

06// more answers

What else Claude can do with superagnt_

Go deeper: Schedules for AI agents · Schedules docs · Pricing

sources
  1. Anthropic: Schedule recurring tasks in Claude Cowork
  2. Claude Code docs: cloud scheduled tasks
  3. Claude Code docs: scheduled tasks in the desktop app
  4. OpenAI: Connecting custom MCP servers in ChatGPT