# Independent proof audit, pass 2

No blocking evidence defect found. One nonblocking hygiene finding: the BEFORE recording opens with a leftover TIFF rejection toast for frames 1–7 (0.000–0.500 s). It disappears at frame 8 (0.583 s), before the native picker result. Clean retained native endpoints are identical. This residue does not obscure the action or affect the conclusion; trim that opening if publishing a polished video.

Scope: read-only proof audit for [#16308](https://git.cafe/pingdotgg/t3code/pulls/16308), web only. BEFORE [c4fa4629eb](https://git.cafe/pingdotgg/t3code/commit/c4fa4629eb645b95aae2beef56c70f705b736049); AFTER [8ea57ac19f](https://git.cafe/pingdotgg/t3code/commit/8ea57ac19f782aeaa82003c5ddff73419f35bdef). No repository changes, checkout, pushes, comments, watches, or live T3 home access. This audits evidence, not an additional full code review.

## Row results

Both revisions: Chromium 148.0.7778.96, DPR 1, 900 px viewport height, reduced motion. S1/S2 are 1440 px light/dark; S3/S4 are 390 px light/dark. All behavior rows are 1440 px light. All 142 public PNGs were independently opened and visually examined, including every original, stability pair, raw difference, amplified difference, threshold image and retained diagnostic. Enlarged native-size composer pairs supplemented overview sheets.

Diff0 below counts any changed pixel in the full unmasked frame. Its linked display is absolute per-channel difference amplified 8×, clamped at 255. DiffT counts pixels whose maximum channel difference exceeds 16. Declared affected regions are unchanged from the public policy; outside them the acceptance limit is 0.1% above 16. All 13 full-frame pairs have ZERO outside-region differences, even at diff0. No extra mask, resizing or interpolation was applied. All recomputed raw/amplified/threshold images exactly match delivered files.

| Row | Status | Before / after (capture SHA prefixes) | Diff0 pixels (%) | DiffT pixels (%) | Interpretation |
|---|---|---|---|---|---|
| S1 | PASS | [Before](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/0dff378c51bcaea0/before-S1.png) `d26a78673d63` / [After](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/0e2e2c564f736898/after-S1.png) `16c2671fb83b` | [55303](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/48a285e19625e090/S1-diff0.png) (4.267207%) | [8321](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/1d63299387c5624f/S1-diffT.png) (0.642052%) | TIFF rejection → generic original chip; independent rebuild exact. |
| S2 | PASS | [Before](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/a8c6dd1c89b56abe/before-S2.png) `fcaa6b9662ad` / [After](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/fe75b77db2f4e3a9/after-S2.png) `59da0e837887` | [55310](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/9e91ab8d688f3c9a/S2-diff0.png) (4.267747%) | [8335](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/44ee188a2bc14c75/S2-diffT.png) (0.643133%) | Readable dark theme; same semantic change. |
| S3 | PASS | [Before](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/e5efd14c3e25077e/before-S3.png) `504aee89c4da` / [After](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/960994b9b3f55a40/after-S3.png) `58db782afa8f` | [42459](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/4b687df909eb0dc9/S3-diff0.png) (12.096581%) | [7611](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/af05ad6acb0ada0e/S3-diffT.png) (2.168376%) | Narrow light composer, filename and send controls visible. |
| S4 | PASS | [Before](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/8c8dd9e73345faaf/before-S4.png) `cd3211c8e2d1` / [After](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/d8f77d0082694ff0/after-S4.png) `b6f5e3f83ab3` | [42466](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/ab39ab3081c47ebe/S4-diff0.png) (12.098575%) | [7622](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/7363e3882f551e71/S4-diffT.png) (2.17151%) | Narrow dark composer, filename and send controls visible. |
| B1native-endpoint | PASS | [Before](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/d0d101fe1039e4f3/before-B1native-endpoint.png) `a05cd669d136` / [After](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/d8c07f7c910f03e4/after-B1native-endpoint.png) `a05cd669d136` | [0](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/7ca681a18e6798e3/B1native-endpoint-diff0.png) (0.0%) | [0](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/2e67c2498e69570c/B1native-endpoint-diffT.png) (0.0%) | Native filechooser path; actual platform MIME empty; unchanged generic result. |
| B1-endpoint | PASS | [Before](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/339bdf88cdeb1bf8/before-B1-endpoint.png) `d65a5034e6c1` / [After](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/bb4b52e828e5370c/after-B1-endpoint.png) `c4dcce19f808` | [6557](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/72d411dcdeedfe1c/B1-endpoint-diff0.png) (0.505941%) | [2518](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/ba0ab6306b0a117f/B1-endpoint-diffT.png) (0.19429%) | TIFF-labelled original becomes file; both normal image controls survive. |
| B2png-endpoint | PASS | [Before](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/ecda09799b131508/before-B2png-endpoint.png) `ac5f17f75fa4` / [After](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/2e1fa22b89f1c4c0/after-B2png-endpoint.png) `9a176c23d6da` | [13773](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/d0dcb5797d48210e/B2png-endpoint-diff0.png) (1.062731%) | [9951](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/43d73278edc24b3f/B2png-endpoint-diffT.png) (0.767824%) | Broken DNG image thumbnail replaced with generic file. |
| B2jpg-endpoint | PASS | [Before](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/c9ee4cdb9e7e65d4/before-B2jpg-endpoint.png) `1ef08a607cb5` / [After](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/e2822e692bc2b1a9/after-B2jpg-endpoint.png) `698f22ed1191` | [13762](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/13f411d689cca3f7/B2jpg-endpoint-diff0.png) (1.061883%) | [9951](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/e5375d468cbcd397/B2jpg-endpoint-diffT.png) (0.767824%) | Broken DNG image thumbnail replaced with generic file. |
| B3-endpoint | PASS | [Before](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/711ff27a822f36a2/before-B3-endpoint.png) `dc700ed9ced0` / [After](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/10e5e86399d3e25a/after-B3-endpoint.png) `a05cd669d136` | [3206](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/071e52bcdebb15e4/B3-endpoint-diff0.png) (0.247377%) | [3163](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/5b6f9bf77c83af96/B3-endpoint-diffT.png) (0.244059%) | Captioned generic-MIME paste claimed by attachment handler; DNG sent. Caption retention is not promised. |
| B4empty-endpoint | PASS | [Before](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/a035454e3d6399e7/before-B4empty-endpoint.png) `7d743079d8bd` / [After](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/52f2d285b828b179/after-B4empty-endpoint.png) `a05cd669d136` | [3](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/fd9e3b81a75e7dde/B4empty-endpoint-diff0.png) (0.000231%) | [0](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/ec5818bae193ae32/B4empty-endpoint-diffT.png) (0.0%) | Empty-MIME drop, reference removal and reattach work. |
| B4mime-endpoint | PASS | [Before](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/f1d173cfb83b94e4/before-B4mime-endpoint.png) `7872b3d23aca` / [After](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/bae22de98de9eecc/after-B4mime-endpoint.png) `006471bbf3a1` | [54198](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/6a2751a86034880b/B4mime-endpoint-diff0.png) (4.181944%) | [6869](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/b03735d8d8d56822/B4mime-endpoint-diffT.png) (0.530015%) | MIME-only clipboard original becomes generic file. |
| B5-marker | PASS | [Before](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/dfec2b729aa71cb6/before-B5-marker.png) `acd581263622` / [After](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/a1db34d8c2af46d8/after-B5-marker.png) `acd581263622` | [0](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/535b1d5dd3b40934/B5-marker-diff0.png) (0.0%) | [0](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/afc9b3a90ba8664f/B5-marker-diffT.png) (0.0%) | Reloaded synthetic legacy generic marker at 100 slots visible. |
| B5-endpoint | PASS | [Before](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/8db9f4d82cb60b7f/before-B5-endpoint.png) `59f193df6d00` / [After](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/5196ff888d2f3d99/after-B5-endpoint.png) `cabd393f64c6` | [12](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/660a9225e377664c/B5-endpoint-diff0.png) (0.000926%) | [0](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/c8b35a243712454d/B5-endpoint-diffT.png) (0.0%) | Reattach replaces target marker, keeps 100 slots, no capacity rejection. |

All six aligned JPEG/PNG controls pass with **zero pixels above 16**; raw diff0 counts are 0, 0, 56, 8, 49 and 5 respectively (maximum channel deltas 0, 0, 10, 6, 5 and 5), consistent with small edge antialiasing differences; the 62×62 source rectangles and their alignment are explicit in each report. B2 removes the prior broken DNG thumbnail, shifting unchanged controls one slot; no scale change.

- B1-JPEG-control: PASS, [amplified diff0](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/99faf81f6e7417ac/B1-JPEG-control-diff0.png), [diffT](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/0b6d7fc548d8fe29/B1-JPEG-control-diffT.png), [source rectangles and counts](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/a9d327c4d47f86d1/B1-JPEG-control-diff.json).
- B1-PNG-control: PASS, [amplified diff0](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/7c7b11b97179bea9/B1-PNG-control-diff0.png), [diffT](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/2a62d705f7bcf52d/B1-PNG-control-diffT.png), [source rectangles and counts](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/c497ff2d24999700/B1-PNG-control-diff.json).
- B2png-JPEG-control: PASS, [amplified diff0](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/fbe2839392ba989d/B2png-JPEG-control-diff0.png), [diffT](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/c0e4f8e545182ceb/B2png-JPEG-control-diffT.png), [source rectangles and counts](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/fc38c4635768f72f/B2png-JPEG-control-diff.json).
- B2png-PNG-control: PASS, [amplified diff0](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/ca36da2244bfdc4b/B2png-PNG-control-diff0.png), [diffT](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/690c64924895ca67/B2png-PNG-control-diffT.png), [source rectangles and counts](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/f2cb98c6fe69d5bd/B2png-PNG-control-diff.json).
- B2jpg-JPEG-control: PASS, [amplified diff0](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/00a7c976227c684e/B2jpg-JPEG-control-diff0.png), [diffT](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/124d8f90cebe30ba/B2jpg-JPEG-control-diffT.png), [source rectangles and counts](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/7eb558439071505a/B2jpg-JPEG-control-diff.json).
- B2jpg-PNG-control: PASS, [amplified diff0](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/a29218205a508c43/B2jpg-PNG-control-diff0.png), [diffT](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/0140d14f29cc5b43/B2jpg-PNG-control-diffT.png), [source rectangles and counts](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/7279dce5ea8bd514/B2jpg-PNG-control-diff.json).

Expected render differences are confined to DNG rejection toast/removal, DNG file reference or broken-thumbnail replacement, and resulting composer reflow. B5 marker state is identical before reattachment; canonicalized MIME does not leave a stale generic-MIME marker. No unintended theme, typography, button, image-control, layout or surrounding-shell change was found. No clipping or ghosted overlapping states were observed in affected controls.

## Full motion review

[BEFORE recording](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/5326267c9ce7b958/before-chaptered.mp4) · [AFTER recording](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/742499509dd2a57b/after-chaptered.mp4)

Independently decoded and visually inspected **every encoded frame**: BEFORE 777 frames / 64.750 s; AFTER 820 frames / 68.333333 s, both 12 fps. Exhaustive composer+toast sheets cover all 1,597 frames without temporal sampling. Native-size enlarged transition sheets and whole-window one-second overview sheets supplement that review. This is an exhaustive-frame visual review, not a structural/endpoint-only check or a claim to have listened to audio narration. Chapter captions and embedded chapter timestamps correspond to the actual picker, send, paste/drop, and reattach actions. ffprobe confirms final embedded chapters match final capture metadata (first chapter normalized to 0).

| Chapter | Before start s | After start s | Visual/action assessment |
|---|---|---|---|
| B1native | 0.000 | 0.000 | Native chooser event followed by generic chip; empty Linux MIME documented. Opening baseline residue noted above. |
| B1 TIFF | 4.776 | 4.695 | Before rejects TIFF; after generic original plus JPEG/PNG controls, upload then send. |
| B2 PNG | 13.650 | 13.645 | Before broken raw thumbnail; after file, normal controls render and send. |
| B2 JPEG | 22.512 | 22.541 | Same separate clean-draft case for JPEG label. |
| B3 captioned paste | 30.727 | 31.169 | Before text path, after DNG reference, actual send. |
| B4 empty drop/remove/reattach | 36.839 | 38.280 | Reference disappears when editor content is replaced; reattach and send follow. This is reference removal, not a claim to click a dedicated remove button. |
| B4 MIME-only paste | 45.281 | 46.679 | Before rejects, after accepts generic original and sends. |
| B5 capacity | 50.808 | 53.279 | Reload, target needs-reattach marker among 100, reattachment replaces it without exceeding capacity. |

Both videos include ordinary transient upload 0%/100% and route/sidebar motion. B5 reload temporarily shows the T3 loading screen (before approximately 51.8–56.2 s, after 54.3–59.0 s), then project hydration; it resolves. Provider “Thinking” after dispatch is not used to claim completed provider execution or DNG comprehension. No stuck upload spinner, stale after-state toast, clipping or ghosting was found. Old chapters JSON and diagnostic video stills are retained provenance, not the authoritative final recording timeline.

## Upload and persistence checks

All **214 public URLs were independently downloaded** with curl and their SHA-256 hashes matched the manifest, not merely their local copies. All 214 local artifacts also match. See public-download-verification.json and manifest-and-secret-check.json in the audit bundle.

The real app wire/upload evidence and isolated SQLite projections were checked; every baseline/after persisted attachment file was independently read and hashed (9 before / 12 after across six messages each). The upload body hashes and actual persisted originals match the fixture. TIFF/PNG/JPEG/octet-stream/empty/MIME-only sent DNG variants remain type:file with canonical image/x-adobe-dng after; ordinary controls remain type:image. Supplemental unknown MIME actual-send/persistence proof and 22 classifier branches pass the supplied verifier. Empty/oversized rejection checks pass. No image transformation is applied to DNG bytes.

DNG fixture SHA-256: `ace962421a72c7f041cd0fbac6cdcf8708c2d0fd8c76b8de51a988609cbc3feb` (950 bytes). JPEG: `2f83f900d23a1d29c79f7a65e05dcefaf9a3a5214aa1214371cb45bb350593c2` (2501 bytes). PNG: `5a8ac452168741f1bc5ac8379d1ecd188f0a18ea3257c19e94afea3f8ba93ebc` (416 bytes).

The full evidence index and sanitized capture trace were read. Public textual artifacts were scanned for auth headers, cookies, bearer/JWT values, credential query parameters and token fields: no secret candidates found. Visual captures contain synthetic prompts/fixtures and local origins, not pairing credentials. Private browser storage, startup pairing URLs and dev logs are excluded from audit uploads.

## Independent real-app rebuild

Downloaded the public capture runner, attachment input helper and exact fixtures; verified their manifest hashes. Adapted the runner only to the audit artifact root, retain S1 light, and return before later capture rows. Started the actual app at the AFTER head in this existing worktree, fresh `/tmp/ev-16308-audit`, with no seeding. No baseline checkout or existing dev-state reuse. Paired a new headed Chromium context on Xvfb; the driver selected the project and staged the TIFF-labelled downloaded DNG through the real composer. The resulting upload body independently matches the DNG fixture and received HTTP 204.

Reproduction commands, executed with workdir `/home/ubuntu/wt/pr-16308`:

```sh
T3CODE_DEV_SEED=0 vp run dev --no-seed --home-dir /tmp/ev-16308-audit > /tmp/to-16308-audit/dev-private.log 2>&1
Xvfb :164 -screen 0 1440x1000x24 -nolisten tcp
node /tmp/to-16308-audit/rebuild.cjs
```

The wrapper reads the freshly generated pairing URL internally from the private dev log; it never publishes the URL. The public bundle carries the wrapper, adapted runner, input helper and fixtures so this is a real app reproduction, not execution of make-diffs.py. Dependencies: existing worktree dependencies, installed Chromium matching 148.0.7778.96, Xvfb. Start server/Xvfb separately and run wrapper after the dev log has its startup pairing URL. Fresh port assignment is read from that URL; no origin is baked in.

**Pixel delta versus delivered after-S1.png: 0 / 1,296,000 pixels (0.000000%); maximum channel delta 0; bounding box null.** The two independent rebuilt frames 500 ms apart are also identical. The rebuilt original, amplified zero-diff, metadata and JSON comparison are public audit artifacts. No scratch was committed; uploaded fixtures and trace satisfy the explicitly authorized reproduction requirement.

## FRICTION

- Python urllib received HTTP 403 from the artifact host; curl fetched all 214 artifacts successfully with matching hashes.
- Linux native file selection reports empty MIME; deterministic labelled File inputs cover TIFF/PNG/JPEG instead of misrepresenting the native MIME.
- Baseline PNG/JPEG-labelled raw data stages broken image thumbnails, rather than always rejecting; the matrix correctly reflects observed behavior.
- Prior pass left continuous review pending; this pass completes exhaustive visual review. Superseded timing files remain retained diagnostics and were not used for final timing.
- A short stale baseline toast at the opening is nonblocking evidence residue, explicitly disclosed above.
- Background shell launch did not persist reliably; retained tool sessions ran the fresh app. Browser closed and audit dev listener owners were stopped after confirming their exact port ownership and worktree cwd. Xvfb session exited; no X164 display lock remains.

## Public audit artifacts

[audit-public-bundle.tar.gz](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/0c5fef01d3ca0b45/audit-public-bundle.tar.gz)
![rebuild-S1.png](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/8626283e9036af0a/rebuild-S1.png)
![reproduction-diff0.png](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/103e424b305f3bd0/reproduction-diff0.png)
[audit-bundle-contents.json](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/48525026a70394ad/audit-bundle-contents.json)
[controls-independent.json](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/c403eabcfcfc5ea6/controls-independent.json)

VERDICT: SUFFICIENT