Check if these issues are valid — if so, understand the root cause of each and fix them. If appropriate, use sub-agents to investigate and fix each issue separately.


<file name="frontend/src/lib/tables.ts">

<violation number="1" location="frontend/src/lib/tables.ts:69">
P2: Duplicate CSV headers will overwrite the first mapping because mapping[header] is assigned for every entry. The CSV import guidance is to drop duplicates (keep the first occurrence), so later duplicates should be ignored instead of replacing the initial mapping.

(Based on your team's feedback about dropping duplicate CSV headers.) [FEEDBACK_USED]</violation>
</file>

<file name="tracecat/tables/importer.py">

<violation number="1" location="tracecat/tables/importer.py:60">
P2: `csv.DictReader` emits a `None` key when a row has more values than headers (e.g., an unquoted comma in a cell). Calling `normalize_csv_header(None)` here will raise `AttributeError` because `str.lstrip()` is called on `None`. Filter out `None` keys before normalizing, e.g.:
```python
normalized_row = {
    normalize_csv_header(key): value
    for key, value in csv_row.items()
    if key is not None
}
```</violation>
</file>

<file name="tracecat/tables/router.py">

<violation number="1" location="tracecat/tables/router.py:745">
P2: Rejecting CSV files with duplicate headers conflicts with the CSV import requirement to drop duplicates and keep the first occurrence. This should discard duplicates instead of raising a validation error.

(Based on your team's feedback about CSV imports handling duplicate headers.) [FEEDBACK_USED]</violation>
</file>

<file name="frontend/src/components/tables/table-import-csv-dialog.tsx">

<violation number="1" location="frontend/src/components/tables/table-import-csv-dialog.tsx:130">
P2: Auto-mapping now fills `columnMapping` with "skip" entries for every header, which makes the form valid even when the user hasn’t mapped any column. This lets submissions through despite the “map at least one column” requirement. Tighten validation (or disable submit) to require at least one non-"skip" mapping.</violation>
</file>

Check if these issues are valid — if so, understand the root cause of each and fix them. If appropriate, use sub-agents to investigate and fix each issue separately.


<file name="tracecat/tables/router.py">

<violation number="1" location="tracecat/tables/router.py:750">
P3: Duplicate header columns should be dropped rather than preserved under synthetic keys; keeping them as hidden columns conflicts with the team’s CSV import guidance and can reintroduce unexpected data in rows.

(Based on your team's feedback about dropping duplicate CSV headers.) [FEEDBACK_USED]</violation>
</file>