> For the complete documentation index, see [llms.txt](https://notara.gitbook.io/notara-docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://notara.gitbook.io/notara-docs/automations/scheduled.md). # Scheduled Tasks Scheduled tasks run the Notara agent on a cron schedule — daily, weekly, monthly, or any custom cadence you configure. ## Creating a Scheduled Task 1. In the dashboard, go to **Automations → Scheduled**. 2. Click **+ New Task**. 3. Fill in the task fields: | Field | Description | | ------------------- | -------------------------------------------------------- | | **Name** | Human-readable label (e.g., "Monday Engineering Digest") | | **Prompt** | The instruction the agent receives when the task runs | | **Channel** | The Slack channel where the output is posted | | **Cron expression** | When the task runs (see below) | 4. Optionally select a **Skill** to activate for this task run. 5. Click **Save** — the task starts in **Active** state and will run at the next scheduled time. ## Cron Syntax Notara uses standard 5-field cron syntax: `minute hour day month weekday`. | Expression | Meaning | | ------------------- | -------------------------------------------------- | | `0 9 * * 1` | Every Monday at 9:00 AM | | `0 9 * * 1-5` | Every weekday at 9:00 AM | | `0 8 * * *` | Every day at 8:00 AM | | `0 9 1 * *` | First of every month at 9:00 AM | | `0 17 * * 5` | Every Friday at 5:00 PM | | `*/30 9-17 * * 1-5` | Every 30 minutes during business hours on weekdays | All times are in UTC. If your team is in a different timezone, adjust accordingly (e.g., 9am EST = `0 14 * * *`). **Helpful tool:** [crontab.guru](https://crontab.guru) is a visual cron expression editor. ## Writing a Good Task Prompt The task prompt should be a clear, specific instruction. The agent runs it without any prior context from a conversation, so it needs to be self-contained. **Good prompt:** ``` Run the growth-weekly-digest skill. Pull the past 7 days of activity and post a structured digest with: (1) MRR and new customers from Stripe, (2) top competitor mentions from Perplexity, (3) key highlights in 3 bullets. Keep the total under 400 words. Format for Slack (bold headers, bullet points). ``` **Less effective prompt:** ``` Do the weekly digest ``` If you have a well-designed skill for the task, reference it in the prompt — skills provide all the formatting and tool instructions so the prompt can be short. ## Managing Tasks ### Pausing a Task Click **Pause** on the task card. The task stops running on schedule but retains its configuration. Click **Resume** to re-enable it. ### Editing a Task Click the task name to open the editor. All fields are editable. Changes take effect at the next scheduled run. Every save creates a **revision** — visible in the task's revision history. You can view or restore any previous configuration. ### Archiving a Task Archiving is the two-step path to deletion: 1. Click **Archive**. The task is disabled immediately and marked as Archived. 2. Archived tasks are visible in the automations list with the Archived state filter. 3. Click **Restore** to bring it back. Note: restored tasks start in the **Paused** state — they don't automatically re-enable. 4. To permanently delete, contact . The two-step flow is intentional — it prevents accidentally deleting an active automation. Archiving also disables it, so there's no risk of a "deleted" task continuing to run. ## Run History Each task maintains a run history log showing: * Run time and trigger type (scheduled vs manual) * Duration * Status (success, failed, truncated) * The channel where output was posted Click any run to see the agent's output for that run. Useful for debugging or verifying that an automation ran as expected. ## System Actor Behavior Scheduled tasks run as the **system actor** — not as a specific team member. Key implications: * **No interactive approval prompts**: write tools execute without requiring an admin to approve via Slack DM. The act of an admin creating and enabling the task is the approval. * **No user-facing gates**: destructive confirmation prompts (the in-channel "Confirm/Cancel" Block Kit message) are not displayed for system actor runs. Admins should only give system-actor tasks access to destructive tools if they're certain of the intended action. * **Audit log**: all tool calls from scheduled tasks are logged with `trigger: scheduled` and the task ID. If you're unsure whether a tool is safe to run unattended, test it with **Run Now** first and review the output before enabling the schedule. ## Common Recipes **Weekly sprint summary (every Monday):** ``` Cron: 0 8 * * 1 Prompt: Use the eng-weekly-digest skill to summarize last week's Linear sprint. Channel: #engineering ``` **Daily MRR pulse (every weekday):** ``` Cron: 0 9 * * 1-5 Prompt: Check Stripe and report today's MRR, new customers in the last 24h, and any failed payments. Keep it under 150 words. Format for Slack. Channel: #growth ``` **Monthly board prep (first of month):** ``` Cron: 0 8 1 * * Prompt: Pull the past month's Stripe data and ChartMogul analytics. Generate a board-ready revenue summary: MRR, growth rate, churn, top accounts. Channel: #leadership ``` --- # 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://notara.gitbook.io/notara-docs/automations/scheduled.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.