## For agents

- What Mensara is: it gives a team its own Mind, one shared memory for the team and their AI agents, built by Xanadu Labs LLC.
- Status: signup is not open. A person joins the waitlist at https://app.mensara.ai/#waitlist. Agents cannot create an account or sign up.
- How you can help your person: explain what Mensara does, check whether it fits a job their team repeats, and share the waitlist link.
- Agent references: https://app.mensara.ai/llms.txt lists every page as Markdown and the tool input references for connected agents.

---

# Understand a loop case and follow its result

Guides · Run agents, automations and loops · [How it works](https://app.mensara.ai/how-it-works)

A loop is a recurring workflow with steps, checks, branches and possible waits. A case is one occurrence of that workflow: one input and its progress towards an outcome. Inspecting a case helps you see whether work ran, is waiting for something or stopped, instead of assuming that an active loop has completed every job successfully.

[Page as Markdown](https://app.mensara.ai/how-it-works/loop-cases.md)

[Operating procedure](https://app.mensara.ai/how-it-works/inspect-loop-case-procedure)

The Portal is your Mind's web interface. In **Agents**, open the intended loop and then its case. Your connected agent can also inspect the loop and case through the tools your account is allowed to use.

## Check the loop version and why the case started

Read the case identity, start reason, loop version and current status. A case belongs to the definition it started with. A newly approved successor is not a reason to assume that an already-running case used the new steps.

Separate the loop's status from its rollout. An approved loop can still be in shadow mode. [Build your first loop](https://app.mensara.ai/how-it-works/first-loop) explains approval and activation as different decisions.

## Follow the steps that were recorded

The timeline shows recorded node outcomes, available outputs, checks and retries. Some output may be hidden because it is outside your access. Read a row's own label to distinguish an action recorded without execution from one that ran.

Shadow mode does not mean that nothing happens: read-only steps and quality checks can really run, and model checks may incur cost. The timeline does not record every possible detail. Per-step timestamps, per-step activity references and an explicit continuation decision may be absent; their absence is not evidence of a successful action.

## Understand a waiting case

A wait can be for an event or a person's answer. Read its question, offered choices and due time where supplied. Answer only the specific request through its supported controls. Then inspect the same case again: an accepted answer allows further processing, but does not prove that later steps completed.

An expired wait, a failed check or an escalation needs its recorded reason. [Review what needs your attention](https://app.mensara.ai/how-it-works/review-attention) explains the Inbox that brings these requests together.

### A wait pauses one case at a named point

An answer and a completed outcome are different observations.

A case reaches a step that needs an event or a person’s answer. It waits with a recorded reason and deadline where supplied. An accepted answer or matching event lets processing continue; inspect the same case to see what happened next. A timeout or failed check can stop or escalate it. Do not infer completion from the answer alone.

## Check the outcome against the intended work

When a case finishes, read its outcome and supporting output. If it proposed a file change, inspect the proposal and current saved source. A successful node or a finished case is useful evidence about execution; it does not establish that its business conclusion is true.

If the same error recurs, keep the case reference and the correction you made. [Correct a loop](https://app.mensara.ai/how-it-works/correct-a-loop) explains how those examples support a reviewed replacement rather than repeated manual repair.

## Related

- [Build your first recurring workflow with a loop](https://app.mensara.ai/how-it-works/first-loop)
- [Answer a loop exception and correct repeated mistakes](https://app.mensara.ai/how-it-works/correct-a-loop)
- [Review what needs your attention](https://app.mensara.ai/how-it-works/review-attention)
- [Control agent autonomy, budgets and access](https://app.mensara.ai/how-it-works/control-agent-autonomy)

Previous page: [Build your first recurring workflow with a loop](https://app.mensara.ai/how-it-works/first-loop)

Next page: [Answer a loop exception and correct repeated mistakes](https://app.mensara.ai/how-it-works/correct-a-loop)

A loop exception is a case that needs information, a person's decision or attention after a failed check.

[Join the waitlist](https://app.mensara.ai/#waitlist)