diff --git a/domain/sap10_calculator/docs/HANDOVER_API_SAMPLE_ACCURACY.md b/domain/sap10_calculator/docs/HANDOVER_API_SAMPLE_ACCURACY.md
index 7a7094c3..afbc4b6f 100644
--- a/domain/sap10_calculator/docs/HANDOVER_API_SAMPLE_ACCURACY.md
+++ b/domain/sap10_calculator/docs/HANDOVER_API_SAMPLE_ACCURACY.md
@@ -4,10 +4,15 @@ Point-in-time note. Start from [`AGENT_GUIDE.md`](AGENT_GUIDE.md) for methodolog
 1e-4 bar, the per-line debugging loop, the section helpers, and the suite command.
 
 - **Branch:** `feature/per-cert-mapper-validation`
-- **HEAD:** `9c0a373f` (S0380.225). Next slice: **S0380.226**.
+- **HEAD:** `c236aa58` (S0380.226). Next slice: **S0380.227**.
 - **Baseline (§4 suite):** `tests/domain/sap10_calculator/ backend/documents_parser/tests/`
-  → green (2395 passed, 1 skipped). Pre-existing out-of-scope failures unchanged
+  → green (2397 passed, 1 skipped). Pre-existing out-of-scope failures unchanged
   (stone-§5.6 in `domain/sap10_ml/tests/`; `test_from_rdsap_schema.py::...test_total_floor_area`).
+- **Toolkit (committed):** `scripts/fetch_2026_epc_sample.py`,
+  `scripts/eval_api_sap_accuracy.py`, `scripts/analyse_api_sap_clusters.py`. The 1,000 cached
+  JSONs live in `/tmp/epc_2026_sample/` (gitignored scratch — re-fetch with the sampler;
+  `EPC_SAMPLE_CACHE` overrides the dir). Re-run the eval after any mapper/calculator change to
+  watch the headline move.
 
 ---
 
@@ -67,10 +72,45 @@ Coverage unblocked **788 → 882 computed (+94)**; one real accuracy bug fixed (
 | S0380.223 | `_part_geometry` early-return key contract (RR KeyError) | 5 |
 | **S0380.224** | **loose-jacket cylinder storage loss (Table 2 Note 1)** — was None'd out → zero loss | **22** (mean err +2.29 → +0.45) |
 | S0380.225 | §10.7 no-water-heating default A-F → 12mm loose jacket | 2 |
+| S0380.226 | Elmhurst "Jacket" cylinder insulation → loose-jacket code 2 (Summary path) | (unblocked case 19) |
 
-**S0380.224 is only DIRECTION-validated** (the 22 certs moved toward lodged + §4/golden stayed
-green) — it has **no worksheet pin on the loose-jacket magnitude**. A worksheet with a
-loose-jacket cylinder would close that (see "What to generate" below).
+Headline at HEAD: **882 / 1000 computed, 41.8% < 0.5** (re-run the eval to refresh).
+
+---
+
+## ★ Active worksheet: simulated case 19 — the electric-storage-heater debug
+
+The user generated `sap worksheets/golden fixture debugging/simulated case 19/`
+(`Summary_001431 (2).pdf` + `P960-0001-001431 - 2026-06-04T174437.228.pdf`), purpose-built to
+hit the #1 cluster. It exercises **electric storage heaters** (SAP code 402, control 2402
+auto-charge, 7-hr off-peak tariff) + a **loose-jacket 210 L cylinder** + **WHS 911** (gas
+boiler for water only) + **room-in-roof gables (Party + Exposed) + an alternative wall +
+exposed floor + electric secondary**.
+
+**S0380.226 unblocked extraction** (the "Jacket" label was raising). Running the Summary path
+through the cascade vs the worksheet (rating block) then exposes the cat-7 cluster bugs — our
+**SAP cont 60.2 vs worksheet ~51.2 (+9, the cluster signature)**:
+
+| line ref | ours | worksheet | gap / cause |
+|---|---|---|---|
+| **cost (255)** | 1482.12 | **1816.58** | **−334 → the primary +9 SAP driver.** Likely the Economy-7 off-peak tariff cost split (SAP 10.2 Table 12a / §10c high-rate vs low-rate). START HERE. |
+| Table 2b TF (53) | 0.54 | **0.60** | we apply the ×0.9 separately-timed multiplier, but the Summary lodges "Separate Time Control: No" → should be 0.60. Check `_table_2b_note_b_multiplier_applies` / the override's `separately_timed_dhw` for a WHS-911 + storage-heater dwelling. |
+| HW fuel (219) | ~3642 | 3188.17 | +454 HW over. Tied to the TF bug + how WHS-911 routes storage vs combi loss (the WaterHeating result showed a spurious `combi_loss` and `solar_storage=0` via the section helper — verify against the FULL cascade, the section helper may not mirror it). |
+| fabric (33) | 305.04 | 304.04 | +1.0 W/K — minor; gable / alternative-wall rounding. Low priority. |
+| CO2 (272) | 3124.67 | 3125.85 | ≈ exact. |
+
+**Debug recipe** (reuse the `/tmp/case19*.py` throwaways or rebuild):
+```python
+pages → ElmhurstSiteNotesExtractor(...).extract() → from_elmhurst_site_notes
+→ cert_to_inputs / cert_to_demand_inputs → calculate_sap_from_inputs
+# section helpers: water_heating_section_from_cert / heat_transmission_section_from_cert
+# CI._cylinder_storage_loss_override(epc, main) returns (57)m directly — useful to bisect.
+```
+The worksheet's rating block is block 1 (UK-avg, region 0); the demand block (postcode) is
+block 2. Pin the rating block for SAP/cost, the demand block for PE/CO2.
+
+S0380.224's loose-jacket magnitude can be **worksheet-pinned at 1e-4 here** once the TF bug is
+fixed (worksheet (51)=0.0330, (52)=0.8298, (53)=0.6000, (55)=3.4531, (56)Jan=107.0456).
 
 ---
 
@@ -103,7 +143,20 @@ is what the strict-raise guard exists to prevent.
 
 ---
 
-## ★ What to generate — the single most productive worksheet
+## ★ Additional worksheets that would help most
+
+Case 19 (above) already covers electric storage heaters + loose-jacket cylinder + RR. The two
+that would add the most NEW coverage:
+1. **An electric ROOM-heater dwelling** (SAP code ~691, control 2601/2602) — the **cat-10
+   cluster (43 certs, worst by mean error 10.26)**, which case 19 does not touch.
+2. **A room-in-roof with a SHELTERED gable and an ADJACENT-TO-HEATED gable** (Table 4 types
+   beyond Party/Exposed) — closes the `gable_wall_type` 2/3 raise (14 certs) and pins the
+   Sheltered (U=ext×R0.5) / Adjacent (U=0) U-values the calculator must add.
+
+The original "design one property" guidance (kept below for reference) is what case 19 was
+built from.
+
+## What to generate — the single most productive worksheet (reference)
 
 Heating is one-per-property, so one worksheet can't cover all four broken heating types. But
 **fabric is independent of heating**, so the highest-ROI single artifact bundles the #1