This is the full end-to-end test: run our model, find what it missed, ask Claude to draw exactly those missing stripes, filter Claude's answers through our sail rules, and measure the COMBINED accuracy — not just Claude's accuracy in isolation. On the photos where our model already struggles, a rescue step meaningfully closes the gap. Projected across the whole 332-photo holdout (where most photos have no miss to begin with), the lift is smaller (about +3.3 points) because misses are a minority of all stripes — but on exactly the scans that need help today, this works.
| Step | Stripes found | Recall |
|---|---|---|
| Our model alone | 193 of 284 | 68.0% |
| + Claude rescue on the misses | +71 of 91 missed | 78.0% of misses recovered |
| Combined final | 264 of 284 | 93.0% |
Recall lift: +25.0 percentage points (36.8% relative improvement)
Cost: 0.24 junk lines added per photo by the rescue step.
This is the hard-photo subset (87 photos where the model missed at least one stripe by construction) — not a random sample of all scans. Whole-holdout projection (recall only, not a direct measurement): not available
Source: 87 hard photos (of an ~88-photo AI-recovery test set), 2026-09-06. Whole-holdout figure is a projection from gate0-ablation-scoreboard.json, not a re-measurement.