Monitor and Govern

Users create and edit automations; you govern them. The Monitor page gives Owners and Contributors a view of every automation users created on this agent — what runs, for whom, how it is going — and, within the permission you have been given, the power to pause, resume, disable or delete.

Navigate to Automations → Monitor in the left sidebar.

Automations users set up on this agent — scheduled tasks that run on their own and deliver the result as a new conversation or a WhatsApp message.

The Monitor never edits an automation. Contributors govern; they do not rewrite. What an automation does stays with the user who asked for it.

Who can open it

Reading users' automations is sensitive: they hold people's standing instructions and phone numbers. So access is a deliberate grant, decided per contributor.

Without permission the Monitor shows a warning in place of the list and requests nothing else, so a contributor still sees that the feature exists and knows whom to ask.

The three permissions

Set on the agent's Agent Contributors page (Organization → Agents → Contributors). See Managing Agent Access.

Deleting is not one of these: only Organization Owners and Admins can delete from the Monitor.

The summary

At the top, six counts, always visible whatever the privacy mode because they carry no content:

Browse and search

Below the counts, filters narrow the list:

The table shows Automation, User, Trigger, Schedule, Action, Status, Last Run, Next Run and Created At. Select a row to open its details.

Automation details

A drawer opens on the right, showing:

Execution history

For every run: Scheduled For, Kind (scheduled, run now, retry, catch-up), Status, whether the Condition held, what each Action did, and how long it took. A failure shows a plain-language reason. See Troubleshooting Runs.

History records how a run went, not what it said. The result itself is in the user's conversation, which you can only read if you may read that user's conversations.

Governance actions

Buttons at the bottom of the drawer, according to your permission and the automation's status:

Disable is stronger than Pause: use Pause for a temporary stop the user may reverse, and Disable when you do not want the user to bring it back.

Privacy Mode

Privacy Mode has priority over contributor permission. When the organization has it on:

• Each automation's content is stored encrypted: the name, the original instruction, the condition, the task, the skills and the WhatsApp number — and the creator's name and email.
• Staff see "Hidden — privacy mode" in their place. The list still shows the schedule, action, status and run times, and the summary counts, because those are operational metadata Qlar needs to run the automation.
• Owners and Contributors cannot read the content, whatever they have been granted. Qlar decrypts it only while running the automation, for the user's own benefit.
• Content is not written to run history, error messages or logs.

Automations are not deleted because of Privacy Mode; they are kept as long as they have future runs, and destroyed with their encryption key when deleted. See Privacy Mode.

Transcripts

Each run happens in a conversation, and the task is sent to the agent as a hidden message. In the CMS conversation pages that message, and the arguments of automation tools, are redacted unless the reader is allowed to view automations. So reading a transcript never becomes a back door to someone's automation.

When the policy changes

The Monitor and the Policy work together. A stricter policy shows up here as automations that were paused or stopped, each with a reason such as "Paused: more active automations than the policy allows" or "Stopped: runs more often than the policy allows".

Next step

Learn to read those reasons in Troubleshooting Runs, or see every field of the page in Help → Monitor.