Rituals

A ritual (also referred to as a schedule) is an automated, recurring execution run of an agent on Rival. You configure a specific prompt and a timing schedule, and the agent executes that task on its own at designated intervals, without requiring a user to manually start a chat session.

Rituals are designed to automate routine, time-sensitive operational tasks, such as generating morning executive briefings, running hourly system checks, or compiling weekly departmental reports.

Creating and configuring a ritual

Setting up a scheduled run is managed directly within an agent’s configuration panel.

1. Access agent settings — open the target agent in your workspace editor and navigate to the Rituals section.

2. Initialize new ritual — click New Ritual to launch the setup panel.

3. Define ritual parameters:

   • Name — a clear, descriptive label identifying the purpose of the scheduled run (for example, Daily Outstanding Invoice Audit).
   • Prompt — the exact operational instruction or context message the agent receives at the start of each execution.
   • Schedule configuration — set the exact execution frequency and timing.

4. Save and activate — save the configuration. Rituals can be paused, edited, re-enabled, or removed at any time.

Timing and schedule controls

Rituals utilize standardized cron-style expressions, offering flexible timing controls to match enterprise operational schedules:

• Flexible cadences — configure runs for specific intervals, such as every weekday at 8:30 AM, at the top of every hour, or on the first day of every month.
• Local time zone alignment — always configure timing relative to your local time zone to avoid unexpected timing offsets caused by standard UTC conversions or seasonal daylight saving shifts.

Execution logs and monitoring

Every scheduled run generates an immutable log entry in the agent’s history tab:

• Run telemetry — displays the exact start time, completion status (Success or Failed), and an output preview.
• Full transcript review — expanding a log entry shows the complete step-by-step execution transcript, including tool calls, connector interactions, and data inputs.
• Troubleshooting — if a scheduled run fails, the log captures specific error messages and stack details, providing a clear starting point for resolution.

For organization-wide activity history, see Audit logs.

Common enterprise use cases

• Daily communication summaries — gathering and categorizing high-priority client emails or Slack updates each morning.
• Hourly infrastructure monitoring — checking system uptime status across connected enterprise tools and sending notifications if anomalies are detected.
• Weekly revenue reports — querying CRM connectors to compile sales pipeline updates for executive review every Friday afternoon.
• Monthly record maintenance — auditing internal document stores or running file cleanup routines at month-end.

Critical interaction with guardrails

Because rituals run autonomously in the background, there is no user active in a chat window to grant real-time permissions:

• Automated skipping of approval steps — any action or tool marked Needs approval is automatically skipped during a ritual run, and an explanatory note is written to the execution log.
• Configuring action permissions — to ensure a ritual completes its tasks without stopping, any system action or tool required for that scheduled run must be explicitly set to Allowed in the agent’s guardrail settings.

Unattended runs skip approvals

A ritual never queues an action for later sign-off. If a step needs human judgment, keep it out of the scheduled prompt and handle it through approval workflows in an interactive session instead.

Enterprise governance and pre-launch checklist

Team members can schedule rituals within their assigned workspace permissions. However, any ritual that interacts with shared enterprise databases, financial tools, or customer communication channels remains subject to global administrative policies and connector rules. See Guardrails and allowed models.

Pre-launch checklist

Before enabling a ritual for production use, complete the following verification steps:

Verification step	Core requirement	Expected outcome
System tools	Verify all required tools and connectors are linked	Agent has valid access credentials
Permission review	Ensure necessary actions are set to Allowed	Zero steps skipped due to missing approval prompts
Manual dry run	Execute the prompt manually in a live chat session	Output format and logic verified before scheduling
Time zone check	Confirm execution timing matches local operational hours	Runs execute at the intended business time

Next steps

• Learn how to set action-level permissions → Personality and guardrails
• Explore full agent architecture → What is a Rival Agent
• Manage administrative oversight across teams → Governance framework