Native evidence: **1/2 rows captured**; synthetic fixture B; before [934c7f8d8e](https://github.com/pingdotgg/t3code/commit/934c7f8d8e7a012aa55be9c0b17498979040fe02), after [0957a043af](https://github.com/pingdotgg/t3code/commit/0957a043af9ef9cd6507964a9533a2b3c7a66e0c).
M2 environment: Pixel 9, Android 16/API 36, 411.43 dp wide (1080×2424 px, density 2.625), font scale 1.0, dark, UTC; client clock frozen at 2026-11-10 06:50 UTC on both sides, status bar hidden identically.
M2 steps, driver elapsed before/after: 0s Cost/30d; 9/11s scroll; 21/25s bottom; 36/47s Tokens/30d; 45/56s 24h; settle. Published clips end at settled Tokens/24h; raw recordings retain the optional final Cost switch.
[Before recording](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/7453136f2fb41eb4/M2-before.mp4) · [After recording](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/99070a7df57c8aec/M2-after.mp4)

| Native row | Before | After | Diff0 | DiffT |
|---|---|---|---|---|
| M2 · Pixel 9 · dark · Cost/30d, top | ![Before](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/15348056bd187f40/M2-before-cost30.png)<br>SHA-256 `9c087bc49283` | ![After](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/7b5cca8f235449b3/M2-after-cost30.png)<br>SHA-256 `3477ced3baec` | ![Zero threshold ×8](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/c533865c769ef859/M2-diffs-cost30-diff0.png)<br>SHA-256 `142d3aa042be` | ![Threshold 16/255](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/8826b5fb9fc3ba64/M2-diffs-cost30-diff16.png)<br>SHA-256 `f7b3a8bc0361` |
| M2 · Pixel 9 · dark · Cost/30d, bottom | ![Before](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/33dde24324d96170/M2-before-model30.png)<br>SHA-256 `f850ad0d1668` | ![After](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/1fa3c4992b83acf7/M2-after-model30.png)<br>SHA-256 `2a4f3ca3b113` | ![Zero threshold ×8](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/b5057688317dedc8/M2-diffs-model30-diff0.png)<br>SHA-256 `3660f213dbe9` | ![Threshold 16/255](https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/0d58e38143f06b8e/M2-diffs-model30-diff16.png)<br>SHA-256 `454f772298ba` |

M1 environment: iPhone 17 Pro lease opened; client installed in 25s; OS, pt width, font scale and requested dark theme could not be verified.
| Native row | Before | After | Diff0 | DiffT |
|---|---|---|---|---|
| M1 · iPhone 17 Pro · dark requested | Missing | Missing | Missing | Missing |
M1 stopped after dark-mode command and one retry >60s later both returned `Error (COMMAND_FAILED): Remote daemon is unavailable`; diagnostics `mwafleoh-98d0b400`, `mwafn4ec-2e9c63bb`. Both leases closed successfully.

- What else changed: cost headline/provider/model $0/Unpriced → $3.42; provider share/chart become positive; cache savings $0 → $0.78 (0.2× raw cost); Unpriced 100% → 0%; Cost type/speed sections appear ($1.71 Standard + $1.71 Fast; $0 premium). Tokens remain 360K/one session; token charts unchanged.
- Capture differences: identical 800 px swipe landed a few pixels differently; bottom-anchored views move earlier content when Cost appears. Top views have identical scroll offset. Diff0 = absolute RGB ×8; DiffT = max channel difference >16/255. All 12 retained frames passed pixel-identical pairs with a 500 ms pause between screenshot commands.
AI-assisted; verified with the tests and tools listed above.