# DefectDojo independent finding-field comparison

> Do the projected finding fields match between two native report importers and the pinned upstream parsers?

Source: https://keryn.dev/evals/defectdojo-fields/

Do the projected finding fields match between two native report importers and the pinned upstream parsers?

| Evaluation detail | Description |
|---|---|
| Keryn engine | Keryn native report importers at `49efc6ec23f4` |
| Reference | Unmodified DefectDojo Django parsers 5.2.16 |
| Corpus | Seven saved JSON reports, including zero-finding controls, restated from `crates/dd/generated/oracle/coverity-tfsec-fields.json` |
| Comparison | Exact equality for the selected field values after normalization, in saved report row order. |
| Receipt | [`defectdojo-fields.json`](https://keryn.dev/evals/defectdojo-fields.json), sha256 `1a6bb22535c2a9b9` |

## Results

| Measure | Value | Count | Denominator |
|---|---|---|---|
| Adapters with matching projections on every compared report | **100%** | 2 / 2 | the two compared adapters |
| Matching normalized field values | **100%** | 336 / 336 | field values in the selected projection on the saved reports |

Cases by status: 2 identical.

## What this does and does not measure

- Only these two adapters and the selected finding-field projection were compared.
- Saved reports were treated as data. Identity and settings for the independent local oracle are pinned in the source receipt.
- Other importer IDs, unprojected fields, application persistence, and unseen report variants remain outside this score.

## All cases (2)

**Show every case**

| Case | Status | Reference rows | Keryn rows | Matching | Missing | Extra |
|---|---|---|---|---|---|---|
| `Coverity` | identical |  |  |  |  |  |
| `TFSec` | identical |  |  |  |  |  |
