Monitor and incidents

The monitor answers two questions: how is each case doing? and what needs someone’s attention? In the Flujos (Flows) capability of the menu, the Casos (Cases) page lists every case with its progress, Incidentes (Incidents) gathers the ones stopped by an error, grouped so you can handle them in one go, and Monitoreo (Monitoring) tells you how your flows are doing. The pages refresh on their own while you are looking.

Cases

Under Centro › Flujos › Casos (Center › Flows › Cases) you see each case with its flow, its status, when it started, its Duración (Duration) and its latest activity.

Status	Meaning
En curso (Running)	The case is moving forward.
En espera (Waiting)	It is waiting on something: a robot, a person’s decision, a scheduled retry, or a time or an event in a wait step. While it only has open waits, it ties up no robots or workers.
Con incidente (Incident)	A step failed and needs someone to decide.
Completado (Completed)	It reached an end (or a person marked it as done).
Cancelado (Cancelled)	It was cancelled by hand or by the system.

Filter by Flujo (flow), Estado (status), Clave de negocio (business key, for example the invoice number) and the Desde / Hasta (from / to) dates.

Duration

The Duración tells you how long each case has been going or how long it took, with no math on your part:

Finished case: the time from start to finish, for example «45 s», «4 min 13 s», «2 h 5 min» or «3 d 4 h».
Open case (running, waiting or with an incident): «En curso · …» (running), counted up to this moment. The clock keeps ticking while you look at the page.

Each step also shows its time: lleva X (X so far) while it is open (the time in its current state) and duró X (took X) once it finished. In the step panel you see, for example, «duró 3 min · terminó hace 2 h» (took 3 min · finished 2 h ago); hovering over a step in the diagram shows its state and how long it took. Steps that take less than a second (start, end, gateways) show no time.

Case detail

Opening a case shows its name (the title or the business key), the flow and the version it runs on, plus its key facts: Inicio (start), Duración (duration), Última actividad (latest activity), Pasos (steps) and Clave de negocio. If the case finished at a named end, the header says so (for example, «Completado «Registrada»», completed at the “Entered” end).

Diagram. That version’s process, colored by each step’s state: green, completed; blue, in progress; amber, waiting for a robot, a person or a retry, or with a business error; red, failed; gray, skipped, cancelled or not reached. The connections the case went through are highlighted, and the business error route it took shows in amber. If something needs attention, a banner at the top takes you to the step.
Step panel. Pick a step to see its state, since when it has been like that or how long it took, the attempt number and, if it failed, what happened and in which robot activity. From there you open Ver la ejecución del robot (the robot run) or Ver la tarea (the task). With permission to see data, you also get the Datos de entrada (input), Salida (output), Resultado (result) and, if one arrived, the Resultado tardío (late result).
Variables. The case data right now. Variables with a padlock are protected: they hold a person’s decision.
Registro del caso (Case log). Everything that happened, in a single timeline (see below).
Eventos del motor (Engine events). Collapsed by default: the engine’s technical history, including variable changes with their Antes and Después (before and after) values.

From the case you jump to related items in other capabilities: Ver el proceso en Robots (see the process in Robots), Ver la credencial en Conexiones (see the credential in Connections), Ver la base de conocimiento (see the knowledge base), Ver la integración en Conexiones (see the integration in Connections) or Ver el disparador en Robots (see the trigger in Robots). Only the links you are allowed to open show up.

Steps that wait

Wait steps show in amber while they wait, and their panel says what they are waiting for and until when:

Esperar (wait): «Esperando hasta date and time» (waiting until…). When it falls due, the step is completed with the time it fell due and the time it moved on.
Esperar evento (wait for event): «Esperando el evento «name»» (waiting for the event…) and, if it has a time limit, until when («vence date and time», due). From there, anyone with permission uses Enviar señal (send signal). Once it arrives, the step shows where it came from (portal, API or robot), when, and which variables it filled; if it expired, that it took the Vencida (expired) exit.

The Registro del caso (case log) records each wait: when it started waiting, when it fell due, which signal arrived and who sent it (the person, the service account or the robot’s execution), or that the time limit ran out with no signal.

When a robot reports a business error

If a robot reports that the case cannot be processed (a blocked supplier, a duplicate invoice) and the step has a business error route, the case does not stop: it takes that route. The detail shows it at a glance:

A banner at the top: «Step» reported a business error → took the error route «name» to «next step», with the robot’s message and the Ver el paso (see the step) button.
The robot step is marked Error de negocio (business error, amber), not completed, and the error point shows Ruta de error tomada (error route taken).

If the step has no error route, the case goes to Con incidente and the suggested action is Decidir: saltar o reintentar (decide: skip or retry).

Case log

The Registro del caso (case log) brings together in a single timeline, newest first, everything that happened to the case. That way you understand a case without opening five screens:

What the flow did and what the people operating it did (retries, skipped steps, corrected variables).
The logs of each robot run, with the activity where each thing happened.
The life of each task: «Tarea creada para X» (task created for X), who took it, what they decided or whether it expired without an answer.
Every call to the AI or to an API, with its duration and, for the AI, the model, the tokens, the cost and the Fuentes (sources) it cited from your knowledge bases.

Each entry shows where it comes from, its level, the time, the step and the attempt. Clicking one highlights its step in the diagram. With permission to see data, you can expand Lo que se envió (what was sent) and the Respuesta (response), and open Ver la ejecución (see the run) or Ver la tarea (see the task). While a robot is running, the log shows En vivo (live); Ver anteriores (see earlier) loads older entries.

With Filtrar (Filter) you narrow it down to what you are looking for: a Paso (step), a Nivel (level: Información, Aviso or Error; info, warning or error), Solo errores (errors only) or an Origen (source: Flujo, Robot, Persona, IA, Integración or Operador; flow, robot, person, AI, integration or operator). Quitar filtros (clear filters) shows everything again.

Fixing a failed step

When a step fails, the case goes to Con incidente and waits for someone to choose what to do. Pick the step in the diagram: the actions show up in its panel.

Action	When to use it
Reintentar (Retry)	The problem is fixed (the system is back, you corrected some data) and the step can be repeated.
Aplicar resultado tardío (Apply late result)	The robot finished after the step was declared failed. Applying its result lets the case continue as if it had arrived on time.
Saltar (Skip)	The case should move on without this step, or end here.
Detener lo que sigue en curso (Stop what is still running)	The step’s robot or task is still alive (for example, the robot did not answer in time). They are stopped so you can retry without two running at once.
Editar variables (Edit variables)	Some data is missing or wrong. Fix it, then retry.
Cancelar caso (Cancel case)	The case should not go on.

Retrying without duplicates

Before repeating a step, Zoan Cloud asks itself whether the step may already have run. If it certainly did not (it was never sent), it simply retries. If it may have run (the robot ended with an error, disconnected halfway, or an API did not answer), the Reintentar: posible duplicado (possible duplicate) prompt asks you to check the target system before confirming with Reintentar de todas formas (retry anyway). For example, if the robot had already entered the invoice in the ERP before failing, repeating it would enter it twice.

If the step’s robot or task is still running, stop it first with Detener lo que sigue en curso.

To make a robot safe to repeat, use its idempotency key. See Robots in a flow.

Skipping a step

In Saltar «step» you choose ¿Por qué salida sigue? (which exit to take) when the step has several, or tick Terminar el caso aquí y darlo por completado (end the case here as completed), which closes the other open steps. If the step was supposed to fill in variables, edit them first so the following steps have what they need.

When you skip a person task, choosing a decision means deciding on the person’s behalf: only someone the task is assigned to can do it, and the decision is stored and protected.

Editing variables

Editar variables is available while the case has an incident. Only the fields you change are saved; to clear a value, tick Quitar el valor (remove the value). Both the old and new values are recorded. Protected variables (people’s decisions) can only be changed by someone with the override permission. After saving, retry the failed step.

Cancelling a case

Cancelar caso stops the open steps, including running robots and tasks, and the case ends as cancelled. It cannot be undone. If the case came from a queue, its item fails.

Incidents

Centro › Flujos › Incidentes (Center › Flows › Incidents) lists every case stopped by an error, grouped by flow, step and reason. Each group shows how many cases it has, since when, and a suggested action:

Suggestion	Meaning
Reintentar (Retry)	The step never ran: it is safe to try again.
Revisar y reintentar (Review and retry)	The step may have run. Check the target system before repeating it so nothing is duplicated.
Aplicar resultado tardío (Apply late result)	The result arrived after the error: review it and apply it to continue.
Corregir y reintentar (Fix and retry)	Some data or setting is missing: fix it (variables or diagram) and try again.
Decidir: saltar o reintentar (Decide: skip or retry)	It is a business error: decide whether the case goes on without this step or you repeat it once it is sorted out.
Detener y reintentar (Stop and retry)	The robot is still running or did not answer: stop it first, then retry.
Actualizar la sesión de robot (Update the robot session)	Windows rejected the account of the robot session (password, locked or expired account). The robot never ran: fix the session with Ir a Sesiones (go to Sessions) and retry with no risk of duplicates.

Wait incidents

Incident	What happened	What to do
`timer_invalid`	Esperar, or the time limit of Esperar evento, takes its date from a variable that has no value or is not a valid date.	Fix the variable with Editar variables (edit variables) and retry the step.
`timer_too_far`	The Esperar date falls more than a year (366 days) out.	Fix the date or the variable and retry, or skip the step.
`message_expired`	Esperar evento reached its time limit with no signal and the step has no Vencida exit.	If the notice already came some other way, skip the step providing the variables; otherwise decide whether the case goes on or is cancelled. So it moves on by itself next time, add the Vencida exit and publish.

From each group you can Abrir el caso más antiguo (open the oldest case) or Ver todos los casos con incidente (see all cases with incidents) for that flow.

Bulk retry

When many cases failed for the same reason (for example, the ERP was down for an hour), retry them all at once:

Tick the groups you want to retry, or Elegir todos (select all).
Click Reintentar lo elegido (retry selected). Up to 100 steps are retried per round; the rest wait for a second round.
Resultado del reintento (retry result) shows, step by step, which ones were retried and which were not, with the reason.

Steps that may already have run are not repeated in the batch: they are set aside for you to confirm with Revisar y confirmar (review and confirm) if you accept a possible duplicate.

Monitoring

Centro › Flujos › Monitoreo (Center › Flows › Monitoring) tells you how your flows are doing so you can decide where to improve: how many cases come in, how many finish well, how long they take and where they get stuck. It counts the cases that started in the period, with their status today.

Filters: Período (period: 7, 30 or 90 days, or Fechas, dates, with Desde and Hasta), Flujo (flow) and Agrupar (group: automatic, by day or by week).
Summary: Casos iniciados (cases started), Terminaron bien (finished well), Duración típica de un caso (typical case duration, and how long 95 out of 100 take to finish), Detenidos por un error (stopped by an error), Espera por personas (waiting on people) and Costo por caso (cost per case).
Detail: the cases coming in per day or week, how they end, ¿Dónde se va el tiempo? (where does the time go?), a table of steps with their typical duration, the worst case, the waiting time, the failures and the Errores de negocio (business errors), ¿Qué falla más? (what fails most?), how long people take to decide and, per flow, the outcome and the AI and robot cost per case.

From there you jump to Ver los casos (see the cases) or Ver incidentes (see incidents).

On other screens

Executions. Robot runs started by a case read Lanzada por el flujo (launched by the flow), with a link to the case.
Tasks. Tasks created by a case read Del flujo «name» (from the flow), with a link to the case for those allowed to see it.
Queues. Items in a dedicated queue show the status of their case.

Next steps

Robots in a flow: how to avoid duplicates when retrying.
Permissions and plans: who can see data, operate and override decisions.