The run workspace

The workspace combines the input and analysis setup, result viewer, and agent conversation. Open a saved run from Assays to continue reviewing the same analysis. On smaller screens, some controls move into menus; use their labels to locate the corresponding action.

Before starting

Confirm the uploaded file and chosen workflow.
Review the sheet, column mapping, plate roles, and units.
Describe the requested calculation and any grouping, controls, or exclusions.
Submit the analysis and inspect the first progress messages for an incorrect input selection.

Progress and questions

The agent conversation records requests, tool activity, questions, and results. Review it when you need to understand what data or method was used. Access to the full live code panel depends on the account entitlement.

If the agent asks a question, answer with the exact sample, sheet, or column names when possible. A waiting analysis may still occupy an active-analysis slot. Submitted answers are included with the run; follow-up messages count as another run against the shared monthly allowance.

Review the result viewer

For runs with multiple fits, use the dataset selector to choose the result you want. The selected fit determines the corresponding parameter panel. Check its dataset label, model, units, and quality flags before copying a value.

Comparison mode overlays compatible curves. Matching labels, units, and scale determine which curves can be combined. A missing curve may have a different coordinate system; inspect its own view before assuming its analysis failed.

See Understanding results for interpretation and Export and reproduce results for the difference between selected-view and whole-run outputs.

Refine or edit an analysis

Use Refine on a completed run to request a specific change against the existing data and context. For example: “Use the Results sheet and report concentrations in nM,” or “Repeat the analysis with the stated blank subtraction.” Review the resulting method and dataset selection after each change.

When editing data and re-fitting, check the changed cells and retain the reason for the change. Refinements can replace the current result. Export a baseline or preserve a separate run before comparing alternatives. A previous-fit restore in the current browser session is not a durable server-side version history.

Refinements and follow-up messages count as another run against the shared monthly allowance when admitted. See Usage and account limits.

Stop and recover

Use Stop while an analysis is active if the task should not continue. Wait for the stopped state. If stopping fails, reopen the run and check whether it has already completed before sending another request.

Stopping does not guarantee that reserved usage is returned. Read the preserved conversation to identify the correction needed before retrying. For a disconnected or apparently stalled run, reload once and inspect the saved state; follow the troubleshooting guide if progress remains unclear.

Duplicate a run

Duplicate creates a new draft from the run setup. Use it to preserve a separate analysis path. Reattach the intended input file and verify the draft settings before starting; duplicating a setup is not a verified copy of a completed result.

Copy a run link

The link action copies the private run URL for bookmarking or reopening. It does not grant collaborator access. To share a reviewed result, export the appropriate output and check its contents before sending it.

Export a completed result

Select the intended fit or comparison view first. Choose an available export format, wait for the download feedback, and open the downloaded file. Check that the figure, tables, and dataset labels match the result you meant to share. Numeric table exports can include results beyond the selected figure.

Before leaving the workspace