> ## Documentation Index
> Fetch the complete documentation index at: https://docs.nokia.atlas.arenaphysica.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Issue Generation

> Generate AI-powered diagnostic issue reports from data files.

## atlas generate-issue

Generate a structured diagnostic issue report by uploading data files and letting Atlas analyze them.

```bash theme={null}
atlas generate-issue
```

Also available as `atlas upload`.

### Interactive flow

The command guides you through each step:

**1. File selection**

Provide a file path, directory, or the name of a previously uploaded file:

* Local file: `~/data/flight_log.bin` or `flight_log.bin`
* Directory: `~/telemetry/mission_1/` (uploads all files inside)
* Previously uploaded: just the filename if Atlas already has it

**2. Conflict handling**

If a file with the same name was uploaded previously:

* Use the existing uploaded file
* Replace with your local version
* Quit and rename

**3. Optional context**

Add up to 3 existing issue IDs to give Atlas context on related prior findings.

**4. Additional instructions**

Provide free-text guidance for the analysis, for example: "focus on power anomalies during motor spin-up."

**5. Confirmation**

Review your configuration before submitting.

### What happens next

1. Files are uploaded to Atlas (if not already present).
2. A `generate_issue` job starts and a link to the job run is displayed.
3. Real-time progress streams to your terminal.
4. When complete, you receive a link to the diagnostic issue report.

Processing time varies by data size, typically 15–40 minutes for large uploads.

### Examples

```bash theme={null}
atlas generate-issue
```

<Tip>
  The CLI remembers previously uploaded files. If you're running repeated analyses on the same dataset, select "use existing" when prompted to skip re-uploading.
</Tip>