diff --git a/Core/Scripts/BLUEPRINT_FORMAT.md b/Core/Scripts/BLUEPRINT_FORMAT.md
index c701a0f..f905988 100644
--- a/Core/Scripts/BLUEPRINT_FORMAT.md
+++ b/Core/Scripts/BLUEPRINT_FORMAT.md
@@ -57,7 +57,7 @@ Every section: `[u32 tag][u64 payload-length in bytes][payload]`.
 | `TOWN` / `0x4E574F54` | `i32 count`, then per town: `f32 X` · `f32 Y` · `u8 tier` · `u8 isHighwayNode` (0/1) | Tier ordinals from `Enums.cs::TownTier`, append-only, range-checked on read. The highway-node flag is what the generator's road topology was built from (v1 dropped it); carried and exposed on the parsed blueprint, consumed by nothing server-side yet. |
 | `TCRV` / `0x56524354` | 50 B base: `u16 curveVersion` · `f32 knot1..knot4` · `f32 spikeMax` · `f32 sea` · `f32 orangeCeil` · `f32 redCeil` · `f32 benchLo` · `f32 benchHi` · `f32 peakCap` · `f32 tailSlope`. **When `curveVersion ≥ 4`, a 36 B modulation extension follows:** `f32 benchAmp` · `f32 plateauAmp` · `f32 shelfSpanMin` · `f32 shelfSpanMax` · `f32 elevFreqIslands` · `f32 strengthFreqIslands` · `i32 benchSeedOffset` · `i32 plateauSeedOffset` · `i32 strengthSeedOffset`. **When `curveVersion ≥ 5`, a 9 B preset extension follows:** `u8 presetId` (1 = compact, 2 = balanced) · `f32 k5` · `f32 k6` — with the four base knot slots this makes the effective curve unambiguous from the record alone (record total 95 B; the length-framed section + version byte keep every layout change safe) | Optional — present iff the height-redistribution curve shaped this blueprint's `HGTS` (config `TerrainCurve: "v3"`); absent = raw legacy profile. **The `u16 curveVersion` selects the field semantics and the unserialized anchor set:** v1 — knots t1..t4, spikeMax = pooled calibration max, benchLo/Hi = the 50 m plateau's lo/hi. v2 — as v1 but spikeMax = the seed's effective raw pre-curve maximum (per-seed spike normalizer). v3 — knots = K1..K4 of six (K5 = 0.930304, K6 = 1.050720 are version constants, not serialized), spikeMax per-seed, benchLo/benchHi = the fixed 100 m / 220 m shelves. **v4 (current)** — as v3, but benchLo/benchHi are the shelf **BASE** anchors and the extension record carries the spatial-modulation parameters: shelf elevations vary ±benchAmp/±plateauAmp and shelf strength blends the output span across `[shelfSpanMin, shelfSpanMax]`, via Simplex fields seeded `resolvedWorldSeed + seedOffset` at `freq/MapSize` (frequencies stated in undulations per island width). Because spikeMax is per-seed (v2+), blueprints are not reproducible from curve constants alone — that is why it is recorded. **Metadata only:** exported heights are already curved; nothing re-applies the map. Calibration provenance: `HeightCurve.cs` header + the task-05/06/07 reports. |
 | `TDTL` / `0x4C544454` | 30 B: `u16 detailVersion` · `f32 reliefAmpM` · `f32 reliefFreqIslands` · `i32 reliefSeedOffset` · `f32 edgeAmpM` · `f32 edgeFreqIslands` · `i32 edgeSeedOffset` · `f32 edgeMaxShiftM` | Optional — present iff the terrain detail passes shaped this blueprint's `HGTS` (config `TerrainDetail: "v1"`, requires the curve). **`detailVersion` selects the body layout and a reader that does not recognise it SKIPS the section** (leaving detail metadata null) rather than misreading a differently shaped payload — the one section whose body is versioned rather than extended, because v1's layout was retired rather than grown. **v1 (retired, never shipped)** — shelf micro-relief + D8 drainage incision, 38 B; the incision produced grid-aligned artifacts and was reverted whole, so the only v1 payloads that exist are in that batch's own tree. **v2 (current)** — shelf micro-relief (±`reliefAmpM` output metres, shelf-ness weighted) + shelf-edge variation: a per-column shift of the curve's shelf/riser knot block K3/K4/K5, drawn from a Simplex field seeded `resolvedWorldSeed + edgeSeedOffset` at `edgeFreqIslands/MapSize`, amplitude ±`edgeAmpM` **metres of INPUT height** — a displacement of the shelf boundary contour, not an elevation change. `edgeAmpM` is recorded **as applied**, after the clamp to `edgeMaxShiftM` (the preset's band-squeeze bound), so the record always describes the terrain rather than the request. **Metadata only** — heights are already detailed. Machinery: `Tools/Scripts/TerrainDetailPass.cs`. |
-| `EROS` / `0x534F5245` | 62 B: `u16 erosionVersion` · `i32 dropletCount` · `i32 lifetime` · `i32 brushRadius` · `i32 seedOffset` · `f32 carveCapM` · `f32 depositCapM` · `f32 seaMarginM` · `f32 inertia` · `f32 capacityFactor` · `f32 minSlopeM` · `f32 erodeRate` · `f32 depositRate` · `f32 evaporation` · `f32 gravity` · `f32 craterExclFactor` | Optional — present iff the droplet hydraulic-erosion pass shaped this blueprint's `HGTS` (config `Erosion: "v1"`, terrain-water tasks 17–18). **`erosionVersion` selects the body layout; an unrecognised version is SKIPPED whole** (erosion metadata left null), same rule as `TDTL`. **v1 (task 17, superseded)** — 58 B, no `depositCapM`; deposition was bilinear over 4 cells and unbounded, which built isolated cones (measured 15.5 m). The only v1 payloads in existence are in that task's batch tree. **v2 (task 18, current)** — deposition is spread over the same cone brush as carving and bounded per cell by `depositCapM` (`<= 0` = unbounded, the reference model). The FOUR governors are `dropletCount`/`lifetime`/`carveCapM`/`depositCapM`; both caps are enforced against one per-cell NET displacement ledger (positive = carved below the height the pass found, negative = built up above it) and asserted on exit. `seaMarginM` is the flood-guard clamp — no cell is carved below sea + margin, and below-sea cells are untouched in BOTH directions, so the rendered coastline cannot move. Droplets are deterministic from `resolvedWorldSeed + seedOffset` (PCG32); no cell within `craterExclFactor × CraterRadius` of the impact centre is modified. Remaining fields are the droplet-model strength dials; slopes/amounts in metres (1 raw = 251 m). All values recorded **as applied** (post config clamping). **Metadata only** — heights are already eroded, and the classify-side sections (`BIOM`/`WBID`/…) never saw the pass by design. Machinery: `Tools/Scripts/HydraulicErosion.cs`. |
+| `EROS` / `0x534F5245` | 67 B: `u16 erosionVersion` · `i32 dropletCount` · `i32 lifetime` · `i32 brushRadius` · `i32 seedOffset` · `f32 carveCapM` · `f32 depositCapM` · `f32 seaMarginM` · `f32 inertia` · `f32 capacityFactor` · `f32 minSlopeM` · `f32 erodeRate` · `f32 depositRate` · `f32 evaporation` · `f32 gravity` · `f32 craterCoreFactor` · `f32 craterFeatherFactor` · `u8 craterMode` | Optional — present iff the droplet hydraulic-erosion pass shaped this blueprint's `HGTS` (config `Erosion: "v1"`, terrain-water tasks 17–19). **`erosionVersion` selects the body layout; an unrecognised version is SKIPPED whole** (erosion metadata left null), same rule as `TDTL`. **v1 (task 17)** — 58 B, no `depositCapM`; deposition was bilinear over 4 cells and unbounded, which built isolated cones (measured 15.5 m). **v2 (task 18)** — 62 B; deposition brush-spread and per-cell bounded. **v3 (task 19, current)** — replaces v2's single `craterExclFactor` with `craterCoreFactor` (the protected strike core), `craterFeatherFactor` and `craterMode` (`0` = full, `1` = feather). Only v2+ payloads exist outside tasks 17–18's own batch trees. The FOUR governors are `dropletCount`/`lifetime`/`carveCapM`/`depositCapM`; both caps are enforced against one per-cell NET displacement ledger (positive = carved below the height the pass found, negative = built up above it) and asserted on exit. `seaMarginM` is the flood-guard clamp — no cell is carved below sea + margin, and below-sea cells are untouched in BOTH directions, so the rendered coastline cannot move **and the crater's flooded bay can be neither carved open nor silted shut regardless of `craterMode`**. Crater radii are FACTORS of `PRMS.CraterRadius`: nothing inside `craterCoreFactor ×` it is modified; `feather` ramps erosion 0→full from there out to `craterFeatherFactor ×` it, `full` applies full strength immediately. Droplets are deterministic from `resolvedWorldSeed + seedOffset` (PCG32). Remaining fields are the droplet-model strength dials; slopes/amounts in metres (1 raw = 251 m). All values recorded **as applied** (post config clamping). **Metadata only** — heights are already eroded, and the classify-side sections (`BIOM`/`WBID`/…) never saw the pass by design. Machinery: `Tools/Scripts/HydraulicErosion.cs`. |
 | `WBID` / `0x44494257` | `MapSize²` × `u16` water-body id, same pixel order as `HGTS` | Optional (absent = no water data, e.g. a legacy re-encode). `0` = no water, `1` = **the** ocean body, `2..N` = lakes. Ids assigned in deterministic scan order (X outer / Y inner, first-encountered pixel), lakes labeled with the **same 4-connectivity as `CalculateTrueOcean`**. Membership is exactly the generator's water classification — the biome grid's Ocean/Lake pixels and this grid's nonzero pixels are the same set **by construction** (shared predicates). Length must equal `2·MapSize²`. |
 | `WBTB` / `0x42544257` | `i32 count`, then per body (20 B): `u16 id` · `u8 type` (0 ocean, 1 lake) · `u8 salinity` (0 fresh, 1 salt) · `f32 surfaceLevel` · `i32 pixelCount` · `f32 centroidX` · `f32 centroidY` | Optional, paired with `WBID`. **`surfaceLevel` is a documented TRANSITIONAL rule:** one flat level per body — `GetSeaLevel` at the body's pixel centroid (ocean: at the map centre) under the still-live latitude field; superseded by the flat-scalar sea model (minted, lands with the coast change set). The field's per-pixel slope is deliberately NOT baked into any section. **Salinity is a provisional default** (ocean salt, lake fresh) — a placeholder for the future fresh/salt irrigation mechanic, not a mechanic. |
 | `WSRF` / `0x46525357` | `MapSize²` × `u16` quantized water-surface elevation, same pixel order | Optional, paired with `WBID`. `0` is the reserved **no-water sentinel**; a real level `L` (raw height units) encodes as `1 + round(L × 32768)` so it can never encode to 0; decode `(q − 1)/32768` (`BlueprintFormat.EncodeWaterLevel`/`DecodeWaterLevel`). Covers `[0 … ~1.99997]` raw at `1/32768` raw ≈ **7.7 mm** of world height (1 raw = 251 m) — far finer than the 1 m voxel. Nonzero exactly where `WBID` is nonzero; the value is the pixel's body level. |
@@ -135,6 +135,15 @@ body).
   Output-height only: the classify path reads pre-erosion heights, so `BIOM` and every water
   section stay bit-identical with erosion on or off, and the sea clamp keeps even the RENDERED
   coastline fixed. When on, `EROS` records the parameters as applied.
+- **`CraterErosionMode`** (`"feather"` | `"full"`, default `"feather"` pending the task-19 gate),
+  with **`CraterErosionCore`** (`0.50`) and **`CraterErosionFeather`** (`1.05`), both factors of
+  `CraterRadius` — how erosion treats the crater surrounds. Task 17's hard `1.2 ×` cutoff left a
+  visible un-eroded disc: the carve writes only inside `0.80 ×` and its displacement is exactly 0
+  beyond that, so 620 811 land cells of ordinary terrain were held smooth for no geometric reason
+  (measured, seed 1280587109). Only the deep strike core is protected now; `feather` ramps erosion
+  in across `core → feather` for a younger-looking crater with no seam, `full` weathers the
+  surrounds like any other terrain. Neither mode can affect the flooded bay or its sea connection —
+  that is the sea clamp's guarantee, not the exclusion's.
 
 - **`CoastProfile`** (`"wide"` | `"steep"`, default `"wide"`) — the submarine shelf. The height
   curve is identity at and below sea level, so it never reshaped the seabed; `"wide"` compresses
diff --git a/Tools/Scripts/README.md b/Tools/Scripts/README.md
index b6e3a60..cf7260a 100644
--- a/Tools/Scripts/README.md
+++ b/Tools/Scripts/README.md
@@ -53,7 +53,18 @@ Generates the entire 2D blueprint. Roughly in order:
    `ErosionDepositCap`; both caps are per-cell metres against one net-displacement ledger and are
    asserted on exit), a sea clamp forbids carving below sea + margin and leaves below-sea cells
    untouched in both directions (the rendered coastline cannot move — asserted every generation),
-   and nothing within 1.2× the crater radius is modified. Output-height only: biome/water
+   and the crater is handled by `CraterErosionMode` (task 19): nothing inside the protected strike
+   core (`CraterErosionCore`, **0.80 ×** CraterRadius — the carve's own extent) is modified, and
+   outside it either `"full"` applies full strength at once or `"feather"` ramps 0→full out to
+   `CraterErosionFeather` (1.05 ×). Task 17's hard 1.2 × cutoff was replaced because it held
+   **620 811 land cells** of ordinary terrain smooth for no geometric reason — measured, the carve's
+   displacement is exactly 0 beyond 0.80 × — which read as an un-eroded disc with a hard edge.
+   Keeping the core at the carve radius is what makes erosion and the carve touch **disjoint**
+   cells; a narrower core lets the carve, which runs afterwards and scales height toward the sea
+   target, amplify erosion's deltas across the waterline (measured at a 0.50 core: 79 rendered
+   waterline crossings). The flooded bay and its sea connection never depend on the crater mode —
+   below-sea cells are read-only in both directions, so the bay can be neither carved open nor
+   silted shut. Output-height only: biome/water
    classification reads the retained pre-erosion map, so `1_biomes`/`0_water` are bit-identical
    with erosion on or off.
    The task-18 defaults tune for a drainage **hierarchy** — fine rills everywhere feeding a set of