> For the complete documentation index, see [llms.txt](https://docs.feathery.io/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.feathery.io/platform/computer-agents/review-a-run.md). # Review a Run Watch an agent work, respond when it needs you, and review what it did. The agent's **Runs** tab lists every run, along with the status, who or what triggered the run, and how long it took. Runs that need your attention sort first, then newest. Click a row to open the run. ## Run Statuses * **Pending**: the run is queued and the browser has not started. * **Running**: the agent is working and can be watched live. * **Action required**: the run is paused, waiting for your approval or an answer. The browser holds its place until you respond or the approval window closes. * **Succeeded**: the agent finished the task and reported an outcome. * **Failed**: the agent could not finish. The run displays the reason. * **Cancelled**: the run was stopped. ## Watch a Run Live While a run is **Running**, the run page streams the agent's browser under **Live view**. The panel beside it shows: * **Task**: the instructions this run received, with any `{{field}}` placeholders already replaced. * **Activity**: the agent's actions, grouped into the phases the agent names as it works. Pauses, refusals, and values the agent could not enter appear as their own entries. Click `Cancel run` to stop a run in progress. While a run is paused for approval, its browser rests on standby and the live stream stops. The page you were on is preserved — click `Wake it to look around` to open the live browser in a new tab and see where the agent left off. ## Respond to Agent Requests A supervised run pauses and shows an **Agent request** card on the run page when the agent needs you. * **Approvals**: the card describes the exact action the agent wants to take, such as submitting a form. Click `Approve` to let it proceed, or `Give instructions` to send the agent directions instead — it will follow them rather than take the action. * **Questions**: the agent may ask for information the task did not provide. Answer by picking from the options it read off the page or typing a value. Every question is optional — the agent is told which ones you skipped and carries on without them. When the agent offers its best guess, it appears as a one-click reply. The card counts down the approval window. If it expires before anyone decides, the run fails. Each decision — who made it and what they said — stays on the run's activity feed. Viewers see the card read-only; deciding requires the Admin or Editor role. ## Review the Outcome When the run finishes, a **Recording** of the full session replaces the live view. It can take a moment to become available. Time spent waiting for an approval is cut from the video, so the recording shows only the agent working. A successful run reports: * **Notes**: a description of what the agent did. * **Confirmation number**: the reference the site produced, if there was one. * **Couldn't enter faithfully**: any value the agent could not enter as provided, because the site rejected it, truncated it, or offered no matching option. --- # Agent Instructions This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com. ## Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter: ``` GET https://docs.feathery.io/platform/computer-agents/review-a-run.md?ask=&goal= ``` `ask` is the immediate question: it should be specific, self-contained, and written in natural language. `goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.