Run agents, automations and loops

Pause, upgrade or remove an installed agent

An installed agent is a program running a defined job on your Mind with its own access, connections and model limits. Managing it means controlling future runs and the version allowed to run. Use this guide when the job changes, an update is offered or a repeated failure needs investigation.

Contents

The Portal is your Mind's web interface. Open Agents, choose the intended installed agent and inspect its identity, version, status, schedule and last run. Use an owner account or an admin whose delegated management permission covers the action.

Suspend future runs before investigating

The agent detail provides a Suspend or Resume control. Read the resulting status after using it. Suspending future work does not reverse an earlier file change, retract a message or establish that an already-started operation was cancelled.

Keep the failed run or proposal reference and inspect its reason. Check the required source access, service connection, model route and budget window. A missing connection needs the right connection, not a blind reinstall.

Review an offered upgrade

Check the offered publisher and version, changed code, access, connections and effects. The supported upgrade route can bind consent to that particular offer. If the offer changes, inspect it again before approving.

A locally edited installation can conflict with a catalog upgrade. Choose deliberately whether to keep the edited copy or replace it through the offered controls; do not hide the conflict by overwriting the local files. A rollback returns to a previously installed approved version where the supported route permits it. It does not undo that version's earlier external effects.

Check a resumed or upgraded agent

Read its installed version and current schedule. Then observe a normal run and inspect the output. A successful management response or a changed version label is not proof that the job ran successfully. If the output is a proposal, check whether it saved or needs review under the Mind's rules.

Use the installation guide when the program has not yet been installed. Use the automation guide or loop guide for those separate workflow lifecycles.

Remove work that is no longer needed

An authorised operator can remove the installed app through its supported lifecycle route. Check the resulting state and any workflows that relied on its actions. Removal does not erase historical records or automatically reverse the work already performed.

If a management response is lost, inspect the same installed identity before trying again. If your client omits the management action, hand the reference to an authorised operator with a compatible client.

Practical guides for people and agents

Start with one useful job

Join the waitlist

Plain text for AI agents. Same content as the human page.

## 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.

---

# Pause, upgrade or remove an installed agent

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

[Page as Markdown](https://app.mensara.ai/how-it-works/manage-an-agent.md)

The Portal is your Mind's web interface. Open **Agents**, choose the intended installed agent and inspect its identity, version, status, schedule and last run. Use an owner account or an admin whose delegated management permission covers the action.

## Suspend future runs before investigating

The agent detail provides a **Suspend** or **Resume** control. Read the resulting status after using it. Suspending future work does not reverse an earlier file change, retract a message or establish that an already-started operation was cancelled.

Keep the failed run or proposal reference and inspect its reason. Check the required source access, service connection, model route and [budget window](https://app.mensara.ai/how-it-works/control-agent-autonomy#understand-the-model-budget-window). A missing connection needs the right connection, not a blind reinstall.

## Review an offered upgrade

## Check a resumed or upgraded agent

Use [the installation guide](https://app.mensara.ai/how-it-works/install-an-agent) when the program has not yet been installed. Use [the automation guide](https://app.mensara.ai/how-it-works/automate-a-task) or [loop guide](https://app.mensara.ai/how-it-works/first-loop) for those separate workflow lifecycles.

## Remove work that is no longer needed

## Related

- [Install an approved agent and check its first result](https://app.mensara.ai/how-it-works/install-an-agent)
- [Control agent autonomy, budgets and access](https://app.mensara.ai/how-it-works/control-agent-autonomy)
- [Understand a loop case and follow its result](https://app.mensara.ai/how-it-works/loop-cases)
- [Review changes and check what was saved](https://app.mensara.ai/how-it-works/changes-approvals-receipts)

Previous page: [Build a governed AI agent](https://app.mensara.ai/how-it-works/build-an-agent)

Next page: [Automate a routine task with installed actions](https://app.mensara.ai/how-it-works/automate-a-task)

An automation starts an ordered sequence of actions when its trigger occurs.

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