> For the complete documentation index, see [llms.txt](https://docs.cherryai.com.cn/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.cherryai.com.cn/docs/en-us/advanced-basic/scheduled-tasks.md). # Scheduled Tasks Scheduled tasks are for [agents](/docs/en-us/cherry-studio/preview/agent.md) provides the ability to run automatically on a schedule, suitable for periodic needs such as "generating a news briefing every day at 9 AM", "summarizing the team's workload every Monday", and "syncing the data dashboard every hour". * **Core idea**: let Cherry Agent automatically execute according to a preset schedule * **Scope**: any task that can be completed by an Agent can be scheduled * **Run logs**: logs are retained for each run and can be viewed at any time > Recommended reading first [Concepts introduction](/docs/en-us/advanced-basic/concepts-101.md) Learn the concepts related to agents. ## Don't want to fill out the form? Let AI create tasks for you If you're not familiar with Cron expressions and various scheduling fields, the simplest way is to directly ask an **agent with autonomous permissions**(such as the built-in Cherry Claw) to describe your needs: > "At 9 AM every morning, use the latest news to make me a 5-point briefing and send the result to my Feishu." The Agent will automatically determine what kind of scheduled task needs to be created and fill in the fields according to your description (schedule type, time, prompt, etc.), only confirming with you if it needs more information mid-way. {% hint style="warning" %} **Core prerequisites**: * Built-in **Cherry Claw** and the agents you create yourself are all in the default state of **“Autonomous mode off + normal mode”**. * **Before using channels (or scheduled tasks), you must manually enter the agent's edit panel and enable “Autonomous mode”**. * After enabling “Autonomous mode”, the underlying tool-call authorization will be fully taken over automatically, and the original “Permission mode” setting will be hidden accordingly, so no further manual configuration is needed. See [agents](/docs/en-us/cherry-studio/preview/agent.md). {% endhint %} If you want to configure manually or understand what each field means, you can continue with the steps below. ## Manual configuration process ### Prerequisites Scheduled tasks depend on Agent execution, so you must first meet the following: 1. **Enabled** [**API server**](/docs/en-us/advanced-basic/api-server.md) 2. **At least one Agent has been configured**, and that Agent has been enabled in the edit panel's “Basic settings” **Autonomous mode**. * *Note: once Autonomous mode is enabled, the original “Permission mode” setting will be automatically hidden, and the underlying permissions will be taken over automatically for full-auto operation.* Agents that do not meet the above conditions **will not** appear in the dropdown when creating a new task. See detailed explanations of the two settings in [agents](/docs/en-us/cherry-studio/preview/agent.md#di-5-bu-tiao-zheng-zhi-neng-ti-de-ti-shi-ci-gong-ju-yu-ji-neng). {% hint style="info" %} The empty-state tip text for scheduled tasks still uses the earlier names "Soul mode / No-permission mode", corresponding respectively to the current **Autonomous mode / full-auto mode**. {% endhint %}

Empty state when no available Agent is configured

### Create a scheduled task Open `Settings → Scheduled tasks`, click the top of the middle column **+ Add** button, and on the right appears **Add task** form:

Add task form

Meaning of each field: | Field | Explanation | | ------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------ | | **Name** | Display name for the task, making it easier to identify later (e.g. "Daily code review") | | **Prompt** | The instruction sent to the Agent when the task runs, describing "what the Agent should do during this run" | | **Schedule type** | `Interval` / `Cron` / `One-time` Choose one of the three | | **Schedule value** | Corresponds to the schedule type: for Interval, enter the number of minutes; for Cron, enter an expression; for One-time, select a specific date and time | | **Timeout** | Maximum duration of a single run (minutes); leave blank for no limit | | **Send to channel** | (Optional) Select a configured [channel](/docs/en-us/advanced-basic/agent-channels.md), and the task result will be automatically pushed to that IM platform | After filling in, click **Save**, the new task will appear in the middle list and immediately wait to be triggered according to the schedule rules. {% hint style="warning" %} If you selected **Send to channel**, you need to ensure that the channel has an available receiving target (Chat ID) on the corresponding platform. The system will warn you when no user under the channel has ever sent a message to the Bot — **Please send a message to the Bot on the corresponding platform first**, to trigger Cherry Studio to record the Chat ID. {% endhint %} ### Schedule type description {% tabs %} {% tab title="Interval (recommended for beginners)" %} Runs in a loop at fixed minute intervals, for example `60` that is, it runs once every hour. {% endtab %} {% tab title="Cron" %} Accepts standard Cron expressions (5 fields):`minute hour day month weekday`. Common examples: | Expression | Meaning | | -------------- | -------------------- | | `0 9 * * *` | Every day at 9:00 AM | | `*/15 * * * *` | Every 15 minutes | | `0 0 * * 1` | Every Monday at 0:00 | | {% endtab %} | | {% tab title="One-time" %} Select a specific date and time; after running once at the specified time, the task is automatically marked as "Completed" and will no longer be triggered. Suitable for one-time scenarios such as "Remind me to submit my weekly report at 5 PM on Friday." {% endtab %} {% endtabs %} ### Manage existing tasks Each task has full management actions in the list: * **Run**: manually trigger once immediately (does not affect subsequent scheduling) * **Pause / Resume**: temporarily stop scheduling without deleting the task, and resume it when needed * **Edit / Delete**: modify task fields or delete it * **Task status**: Active / Paused / Completed (only one-time tasks will enter "Completed" after the scheduled time) ### View run results Each task retains **run history**: * Click the task card to view past run records: time, duration, status (success / error), output * Failed runs include error information for easier troubleshooting * You can click **View session** to open the full conversation record corresponding to that task run ### Integration with channels The Agent output of scheduled tasks can be [channel](/docs/en-us/advanced-basic/agent-channels.md) automatically pushed to: * Feishu / Lark * Telegram / Discord / Slack * QQ / WeChat For example, "Push the task results to the Feishu group every day at 9 AM." {% hint style="warning" %} Long-running scheduled tasks will continuously consume model tokens. Please set a monthly limit in the Provider dashboard to avoid overspending. {% endhint %} ### Tips and tricks * It is recommended to start with **Interval = 5 minutes, max runs = 1** for a trial run; once the output meets expectations, increase the interval * Different tasks can share the same Agent, or you can prepare dedicated Agents for different scenarios * You can reference in the prompt `{date}` and other placeholders (specific support depends on the Agent template) *** ### 💡 Get help and submit feedback If you encounter any questions, bugs, or have suggestions for feature improvements during configuration or use, please refer to [Feedback and Suggestions](/docs/en-us/question-contact/suggestions.md) the official channels provided there. --- # Agent Instructions This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com. ## Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter: ``` GET https://docs.cherryai.com.cn/docs/en-us/advanced-basic/scheduled-tasks.md?ask=&goal= ``` `ask` is the immediate question: it should be specific, self-contained, and written in natural language. `goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.