CSV Test Import

Use this skill when a user wants to import test cases from CSV into qaos. CSV formats can be arbitrary, so first infer structure and mapping, then run the import script.

Trigger Conditions

• •User asks to import or migrate test cases from a .csv file.
• •CSV has unknown or inconsistent schema.
• •User wants structured PostgreSQL test records.

Instructions

1. •Inspect CSV headers and representative rows.
2. •Infer a mapping from raw CSV columns to normalized keys.
3. •Call the script with run_script using transform-csv.
4. •Report import summary (insertedCount, skippedCount) and any skipped-row reasons.

run_script template

{
  "action": "run_script",
  "skill_name": "csv-test-import",
  "script": "transform-csv",
  "args": {
    "csvPath": "/absolute/path/to/file.csv",
    "testId": "import-2026-02-21",
    "mapping": {
      "Test case": "title",
      "Priority": "priority",
      "Description": "description",
      "Tested by": "testedBy",
      "Test result": "testResult",
      "Bug ticket": "bugTicket"
    },
    "batchSize": 500
  }
}

Notes

• •The script uses stream-based processing (fs.createReadStream) and does not load full CSV into memory.
• •Use larger batchSize for faster imports on stable DB/network.
• •If CSV does not have headers, first derive synthetic headers and adjust mapping accordingly.