diff --git a/scripts/build-kairic-edge-gfx1151.sh b/scripts/build-kairic-edge-gfx1151.sh
new file mode 100755
index 0000000..95390ed
--- /dev/null
+++ b/scripts/build-kairic-edge-gfx1151.sh
@@ -0,0 +1,96 @@
+#!/usr/bin/env bash
+set -euo pipefail
+
+readonly repo_root="$(cd -- "$(dirname -- "${BASH_SOURCE[0]}")/.." && pwd)"
+readonly ck_root="${KAIRIC_CK_ROOT:-$repo_root/third_party/composable_kernel}"
+readonly ck_commit=fdf4bb7fcc984811cef48ce817d89aac064b984a
+readonly ck_patch="$repo_root/patches/composable-kernel-gfx1151-iu4.patch"
+readonly build_dir="${BUILD_DIR:-$repo_root/build-kairic}"
+readonly rocm="${ROCM_PATH:-/opt/rocm}"
+readonly jobs="${JOBS:-$(nproc)}"
+
+for tool in cmake git ninja; do
+  command -v "$tool" >/dev/null || {
+    echo "missing required tool: $tool" >&2
+    exit 1
+  }
+done
+
+[[ -f "$ck_patch" ]] || {
+  echo "missing Composable Kernel patch: $ck_patch" >&2
+  exit 1
+}
+
+if [[ ! -d "$ck_root/.git" ]]; then
+  mkdir -p "$(dirname -- "$ck_root")"
+  git clone https://github.com/ROCm/composable_kernel.git "$ck_root"
+fi
+
+git -C "$ck_root" checkout "$ck_commit"
+if git -C "$ck_root" diff --quiet; then
+  git -C "$ck_root" apply --check "$ck_patch"
+  git -C "$ck_root" apply "$ck_patch"
+elif git -C "$ck_root" apply --reverse --check "$ck_patch"; then
+  echo "Composable Kernel release patch is already applied"
+else
+  echo "Composable Kernel checkout does not match the pinned patch" >&2
+  exit 1
+fi
+
+if [[ -x "$rocm/bin/amdclang++" ]]; then
+  hip_compiler="$rocm/bin/amdclang++"
+elif [[ -x "$rocm/llvm/bin/clang++" ]]; then
+  hip_compiler="$rocm/llvm/bin/clang++"
+else
+  echo "cannot find a HIP compiler below ROCM_PATH=$rocm" >&2
+  exit 1
+fi
+
+if [[ -n "${CC:-}" ]]; then
+  c_compiler="$(command -v "$CC")"
+elif [[ -x /usr/bin/gcc ]]; then
+  c_compiler=/usr/bin/gcc
+else
+  c_compiler="$(command -v gcc)"
+fi
+if [[ -n "${CXX:-}" ]]; then
+  cxx_compiler="$(command -v "$CXX")"
+elif [[ -x /usr/bin/g++ ]]; then
+  cxx_compiler=/usr/bin/g++
+else
+  cxx_compiler="$(command -v g++)"
+fi
+
+export ROCM_PATH="$rocm"
+export PATH="$rocm/bin:$rocm/lib/llvm/bin:$rocm/llvm/bin:$PATH"
+export LD_LIBRARY_PATH="$rocm/lib:$rocm/lib/rocm_sysdeps/lib:$rocm/lib/llvm/lib:$rocm/llvm/lib${LD_LIBRARY_PATH:+:$LD_LIBRARY_PATH}"
+readonly hip_flags="-DGGML_ROCMFPX_RDNA35_MMID_MAX_BATCH=5 -DGGML_ROCMFPX_MOE_MMVQ_ROWS_PER_BLOCK=4${KAIRIC_HIP_TOOLCHAIN_FLAGS:+ $KAIRIC_HIP_TOOLCHAIN_FLAGS}"
+
+cmake -S "$repo_root" -B "$build_dir" -G Ninja \
+  -DCMAKE_BUILD_TYPE=Release \
+  -DCMAKE_C_COMPILER="$c_compiler" \
+  -DCMAKE_CXX_COMPILER="$cxx_compiler" \
+  -DCMAKE_HIP_COMPILER="$hip_compiler" \
+  -DCMAKE_HIP_FLAGS="$hip_flags" \
+  -DCMAKE_PREFIX_PATH="$rocm" \
+  -DBUILD_SHARED_LIBS=ON \
+  -DGGML_CPU=ON -DGGML_OPENMP=ON -DGGML_HIP=ON -DGGML_CUDA=OFF \
+  -DGGML_VULKAN=OFF -DGGML_HIP_FORCE_MMQ=ON -DGGML_HIP_GRAPHS=ON \
+  -DGGML_HIP_MMQ_MFMA=ON -DGGML_HIP_NO_VMM=ON \
+  -DGGML_HIP_ROCWMMA_FATTN=OFF -DGGML_NATIVE=ON \
+  -DAMDGPU_TARGETS=gfx1151 -DGPU_BUILD_TARGETS=gfx1151 \
+  -DPROMPTFORGE_CK_ROOT="$ck_root" \
+  -DLLAMA_BUILD_WEBUI=OFF \
+  -DGGML_BUILD_TESTS=OFF -DLLAMA_BUILD_TESTS=OFF -DLLAMA_BUILD_SERVER=ON
+
+cmake --build "$build_dir" --target llama-server -j"$jobs"
+
+"$build_dir/bin/llama-server" --version
+"$build_dir/bin/llama-server" --help | grep -A1 -- '--kairic-edge'
+ldd "$build_dir/bin/llama-server" 2>&1 | tee "$build_dir/llama-server.ldd.txt"
+if grep -Eiq 'not found|version .* required by .* not found' "$build_dir/llama-server.ldd.txt"; then
+  echo "runtime dependency verification failed" >&2
+  exit 1
+fi
+
+echo "Kairic Edge runtime built at $build_dir/bin/llama-server"
diff --git a/scripts/deploy-remote.sh b/scripts/deploy-remote.sh
index befe272..04623ef 100755
--- a/scripts/deploy-remote.sh
+++ b/scripts/deploy-remote.sh
@@ -1072,20 +1072,21 @@ deploy_anubis() {
     wait_container_healthy anubis 60 || return 1
 }
 
-# frank-large is the unsloth llama.cpp prebuilt (b11030-mix-5ff778e, ROCm
-# gfx1151) serving unsloth/Qwen3.8-Flash-Next-GGUF:UD-IQ4_XS on the Strix
-# Halo iGPU, as a ROOT unit since the 2026-09-19 cutover: the runtime is a
-# prebuilt tarball with its own bundled ROCm, so there is no container
-# image to install and no generated unit to dodge. relink_unit points
+# frank-large is the Kairic Edge runtime (ciru-ai/ROCmFPX, branch
+# kairic-edge-qwen38-27b-v1.2, commit 205a3e5) serving
+# jcbtc/Qwen3.8-27B-IU4-Kairic-Edge on the Strix Halo iGPU, as a ROOT
+# unit since the 2026-09-19 cutover: the runtime is built from source
+# (scripts/build-kairic-edge-gfx1151.sh), so there is no container image
+# to install and no generated unit to dodge. relink_unit points
 # /etc/systemd/system/frank-large.service at the repo file and boot
 # persistence is that unit's [Install] section.
 #
-# It replaces the CIRU v4.4.1 Orca runtime (jcbtc/Qwen3.8-Flash-CIRU-
-# STRIX-Orca), which had replaced the halogen-flash-server quadlet. The
-# port gate is the same as every cutover since 2026-09-16: nothing starts
+# It replaces the unsloth llama.cpp prebuilt (b11030-mix-5ff778e, ROCm
+# gfx1151), which had replaced the CIRU v4.4.1 Orca runtime. The port
+# gate is the same as every cutover since 2026-09-16: nothing starts
 # until 7770 is released.
-LLM_LARGE_MODEL="${LLM_LARGE_MODEL:-/home/dank/models/Qwen3.8-Flash-Next-UD-IQ4_XS/UD-IQ4_XS/Qwen3.8-Flash-Next-UD-IQ4_XS-00001-of-00003.gguf}"
-LLM_LARGE_DRAFT="${LLM_LARGE_DRAFT:-/home/dank/models/Qwen3.8-Flash-Next-UD-IQ4_XS/mtp/mtp-Qwen3.8-Flash-Next-shared-Q8_0.gguf}"
+LLM_LARGE_MODEL="${LLM_LARGE_MODEL:-/home/dank/models/Qwen3.8-27B-IU4-Kairic-Edge/Qwen3.8-27B-IU4-Kairic-Edge.gguf}"
+LLM_LARGE_DRAFT="${LLM_LARGE_DRAFT:-/home/dank/models/Qwen3.8-27B-IU4-Kairic-Edge/Qwen3.8-27B-Kairic-IU4-FFN.pfs}"
 
 retire_halogen_large_quadlet() {
     local trash="$HOME/.AGENTTRASH/20260919-orca-cutover/quadlet"
@@ -1125,7 +1126,7 @@ deploy_llm_large() {
     for f in "$LLM_LARGE_MODEL" "$LLM_LARGE_DRAFT"; do
         if [[ ! -f "$f" ]]; then
             echo "ERROR: payload absent at $f" >&2
-            echo "frank-large: fetch it: bash scripts/download-frank-large-unsloth.sh" >&2
+            echo "frank-large: fetch it: bash scripts/download-frank-large-kairic.sh" >&2
             missing=1
         fi
     done
@@ -1140,7 +1141,7 @@ deploy_llm_large() {
     # The pattern never appears in the deploying shell's own command line, so
     # -f is safe here (it is not in `bash deploy-remote.sh`).
     run_privileged systemctl stop frank-large.service 2>/dev/null || true
-    pkill -f 'Qwen3.8-Flash-Next-UD-IQ4_XS' 2>/dev/null || true
+    pkill -f 'Qwen3.8-27B-IU4-Kairic-Edge' 2>/dev/null || true
     # Nothing may start until the predecessor has released the port. On
     # 2026-09-16 the cutover failed exactly here: the old server logged
     # "Deactivated successfully", the next engine tried to bind 7770 a second
diff --git a/scripts/download-frank-large-kairic.sh b/scripts/download-frank-large-kairic.sh
new file mode 100755
index 0000000..a730de3
--- /dev/null
+++ b/scripts/download-frank-large-kairic.sh
@@ -0,0 +1,110 @@
+#!/usr/bin/env bash
+# Download jcbtc/Qwen3.8-27B-IU4-Kairic-Edge (the frank-large model) on
+# Dankclaw.
+#
+#   bash scripts/download-frank-large-kairic.sh
+#   bash scripts/download-frank-large-kairic.sh --verify-only
+#
+# The Kairic Edge payload is four files: the main IU4 GGUF (15.5 GiB) and
+# three .pfs sidecars (FFN 7.99 GiB, GDN 1.88 GiB, GDN-Output 0.70 GiB).
+# The sidecars are the PromptForge IU4 (W4A4) weight shards the Kairic Edge
+# runtime loads alongside the main GGUF (PROMPTFORGE_IU4_SIDECAR,
+# PROMPTFORGE_GDN_SIDECAR, PROMPTFORGE_GDN_OUTPUT_SIDECAR).
+#
+# Integrity: every file is checked against its declared byte count and sha256
+# (the values below are the LFS oids from the Hub, 2026-09-21). The repo is
+# public (gated: false); an HF token at ~/.cache/huggingface/token is used
+# when present to clear the anonymous rate limit on a 26 GB transfer.
+#
+# Layout after a successful run (MODEL_DIR):
+#   Qwen3.8-27B-IU4-Kairic-Edge.gguf        16617792672  main IU4 GGUF
+#   Qwen3.8-27B-Kairic-IU4-FFN.pfs          8576856064  FFN sidecar
+#   Qwen3.8-27B-Kairic-IU4-GDN.pfs          2019569664  GDN sidecar
+#   Qwen3.8-27B-Kairic-IU4-GDN-Output.pfs    756953088  GDN-Output sidecar
+set -euo pipefail
+
+REPO="jcbtc/Qwen3.8-27B-IU4-Kairic-Edge"
+MODEL_DIR="/home/dank/models/Qwen3.8-27B-IU4-Kairic-Edge"
+BASE="https://huggingface.co/$REPO/resolve/main"
+
+mode="download"
+[[ "${1:-}" == "--verify-only" ]] && mode="verify"
+
+token=""
+auth=()
+if [[ -r "$HOME/.cache/huggingface/token" ]]; then
+    token="$(tr -d '[:space:]' < "$HOME/.cache/huggingface/token")"
+    auth=(-H "Authorization: Bearer $token")
+fi
+
+dl() { # dl <relpath>
+    local rel="$1" dest="$MODEL_DIR/$rel"
+    [[ -s "$dest" ]] && return 0
+    mkdir -p "$(dirname "$dest")"
+    echo "fetching $rel"
+    curl -fL -C - "${auth[@]}" --retry 5 --retry-delay 3 -o "$dest" "$BASE/$rel"
+}
+
+# filename size sha256 — the LFS oids from the Hub (2026-09-21)
+FILES=(
+    "Qwen3.8-27B-IU4-Kairic-Edge.gguf 16617792672 360caf7381907c3eca7ac0afd1228efc016af747f3f38637fb1c7f94daabac2a"
+    "Qwen3.8-27B-Kairic-IU4-FFN.pfs 8576856064 adcbb90a7b429a30a2a39043366d68320d72e8b4816a0f498e882b2f80a2ba2b"
+    "Qwen3.8-27B-Kairic-IU4-GDN.pfs 2019569664 82f931316f1c895da104915dec4697163808d06f0e6b2dc027cee7aa3afc0f0e"
+    "Qwen3.8-27B-Kairic-IU4-GDN-Output.pfs 756953088 3b07e7b176559e4402924ba0c368532fa6f02118a33c71e70974c809bf6208a3"
+)
+
+RELS=()
+for line in "${FILES[@]}"; do
+    read -r rel _ _ <<<"$line"
+    RELS+=("$rel")
+done
+
+# Verify each file's byte count and sha256 (idempotent: a complete file is
+# re-checked, not re-downloaded).
+echo "verifying $MODEL_DIR against declared sizes + sha256"
+python3 - "$MODEL_DIR" "${RELS[@]}" <<'PY'
+import hashlib, os, sys
+model_dir = sys.argv[1]
+rels = sys.argv[2:]
+# (rel, size, sha256) triples, re-parsed from the same source of truth.
+specs = {
+    "Qwen3.8-27B-IU4-Kairic-Edge.gguf": (16617792672, "360caf7381907c3eca7ac0afd1228efc016af747f3f38637fb1c7f94daabac2a"),
+    "Qwen3.8-27B-Kairic-IU4-FFN.pfs": (8576856064, "adcbb90a7b429a30a2a39043366d68320d72e8b4816a0f498e882b2f80a2ba2b"),
+    "Qwen3.8-27B-Kairic-IU4-GDN.pfs": (2019569664, "82f931316f1c895da104915dec4697163808d06f0e6b2dc027cee7aa3afc0f0e"),
+    "Qwen3.8-27B-Kairic-IU4-GDN-Output.pfs": (756953088, "3b07e7b176559e4402924ba0c368532fa6f02118a33c71e70974c809bf6208a3"),
+}
+ok = True
+for rel in rels:
+    dest = os.path.join(model_dir, rel)
+    size, sha = specs[rel]
+    if not os.path.isfile(dest):
+        print(f"  MISSING {rel}")
+        ok = False
+        continue
+    actual_size = os.path.getsize(dest)
+    if actual_size != size:
+        print(f"  SIZE MISMATCH {rel}: {actual_size} != {size}")
+        ok = False
+        continue
+    h = hashlib.sha256()
+    with open(dest, "rb") as f:
+        for chunk in iter(lambda: f.read(1 << 24), b""):
+            h.update(chunk)
+    actual_sha = h.hexdigest()
+    if actual_sha != sha:
+        print(f"  SHA256 MISMATCH {rel}: {actual_sha}")
+        ok = False
+        continue
+    print(f"  OK {rel} ({actual_size} bytes)")
+if not ok:
+    sys.exit(1)
+PY
+
+if [[ "$mode" == "download" ]]; then
+    for line in "${FILES[@]}"; do
+        read -r rel _ _ <<<"$line"
+        dl "$rel"
+    done
+fi
+
+echo "frank-large payload complete: $MODEL_DIR"
diff --git a/scripts/frank-large.env b/scripts/frank-large.env
index 203dafb..a6ec77a 100644
--- a/scripts/frank-large.env
+++ b/scripts/frank-large.env
@@ -1,6 +1,6 @@
-# frank-large — unsloth llama.cpp prebuilt (b11030-mix-5ff778e, ROCm
-# gfx1151) serving unsloth/Qwen3.8-Flash-Next-GGUF:UD-IQ4_XS on the Strix
-# Halo iGPU.
+# frank-large — Kairic Edge runtime (ciru-ai/ROCmFPX, branch
+# kairic-edge-qwen38-27b-v1.2, commit 205a3e5) serving
+# jcbtc/Qwen3.8-27B-IU4-Kairic-Edge on the Strix Halo iGPU (gfx1151).
 #
 # These are the tunables read by scripts/serve-frank-large.sh. The flag
 # list is owned by the serve script; this file only sets fleet values.
@@ -9,31 +9,34 @@
 # runtime reads those bare (the 2026-08-13 TEMP incident). HOST is the one
 # exception because the serve script itself consumes it for --host.
 
-# The assembled payload (scripts/download-frank-large-unsloth.sh): the
-# serve script refuses to start without the main GGUF and the MTP head.
-MODEL_DIR=/home/dank/models/Qwen3.8-Flash-Next-UD-IQ4_XS
+# The assembled payload (scripts/download-frank-large-kairic.sh): the
+# serve script refuses to start without the main GGUF and the three
+# sidecars (FFN, GDN, GDN-Output).
+MODEL_DIR=/home/dank/models/Qwen3.8-27B-IU4-Kairic-Edge
 
-# The prebuilt runtime (see the serve script for the fetch line).
-RUNTIME_DIR=/home/dank/GitPackages/unsloth-llama-b11030
+# The Kairic Edge runtime (scripts/build-kairic-edge-gfx1151.sh).
+RUNTIME_DIR=/home/dank/GitPackages/ROCmFPX-kairic/build-kairic
 
-# Fleet endpoint: the same port and alias the CIRU predecessor served, so
+# Fleet endpoint: the same port and alias the unsloth predecessor served, so
 # no client, probe or models.yml entry changes address.
 HOST=0.0.0.0
 PORT=7770
 
-# One stream, 200k context (operator's call, 2026-09-21). The model's
-# trained window is 262144; 200000 is the requested serving shape.
-CONTEXT_SIZE=200000
+# One stream, full 262144 context (the model's native window).
+CONTEXT_SIZE=262144
 PARALLEL_SLOTS=1
 
-# MTP shared-Q8_0 head: on. Draft depth 2 is the model card's default
-# ("higher drafts more but each guess is accepted less often"). The shared
-# head borrows the token embedding and output projection from the main
-# model, so the file is 2.6 GB rather than 4.1.
-ENABLE_MTP=1
-SPEC_DRAFT_N_MAX=2
+# Native MTP depth 4 (the model's MTP layers) + n-gram drafting (24/64/64).
+# DFlash is NOT available for this model (no DRAFT_DFLASH spec type in the
+# Kairic Edge runtime).
+SPEC_DRAFT_N_MAX=4
 
-# Text-only for now: the F16 projector is on disk at vision/mmproj-F16.gguf;
-# flip to 1 to enable it. The fleet's image path is frank-small either way,
-# so models.yml stays input: [text] until this changes.
-ENABLE_VISION=0
+# The harness needs sampling, tools and a reasoning budget, which the
+# fast-greedy path rejects. COMPATIBILITY_MODE=1 enables them (9.7% slower
+# per the build guide).
+COMPATIBILITY_MODE=1
+
+# Prompt-cache budget: 8 GiB (the Kairic Edge qualified config). The model
+# is 27B dense (not DeltaNet), so the fleet's 12.8 GiB DeltaNet floor does
+# not apply.
+CACHE_RAM=8192
diff --git a/scripts/frank-large.service b/scripts/frank-large.service
index 4e08322..adf2b30 100644
--- a/scripts/frank-large.service
+++ b/scripts/frank-large.service
@@ -1,5 +1,5 @@
 [Unit]
-Description=unsloth llama.cpp (b11030, ROCm gfx1151) Qwen3.8-Flash-Next UD-IQ4_XS (dankclaw, frank-large, Strix Halo iGPU)
+Description=Kairic Edge (ROCmFPX v1.2, gfx1151) Qwen3.8-27B IU4 W4A4 (dankclaw, frank-large, Strix Halo iGPU)
 After=network-online.target
 Wants=network-online.target
 
diff --git a/scripts/serve-frank-large.sh b/scripts/serve-frank-large.sh
index 0957758..2b18ac5 100755
--- a/scripts/serve-frank-large.sh
+++ b/scripts/serve-frank-large.sh
@@ -1,21 +1,34 @@
 #!/usr/bin/env bash
-# Serve unsloth/Qwen3.8-Flash-Next-GGUF:UD-IQ4_XS as frank-large on
-# Dankclaw's Strix Halo iGPU (gfx1151), using the unsloth llama.cpp
-# prebuilt (b11030-mix-5ff778e, ROCm gfx1151) — the first runtime with an
-# MTP graph for the qwen4exp architecture (stock ggml-org mainline has no
-# MTP graph, no cross-model tensor borrowing and no --spec-type draft-mtp
-# for this model).
+# Serve jcbtc/Qwen3.8-27B-IU4-Kairic-Edge as frank-large on
+# Dankclaw's Strix Halo iGPU (gfx1151), using the Kairic Edge runtime
+# (ciru-ai/ROCmFPX, branch kairic-edge-qwen38-27b-v1.2, commit 205a3e5) —
+# the qualified IU4 (W4A4) serving path with the FFN, GDN and GDN-Output
+# sidecars.
 #
-# Runtime: the prebuilt app tarball from the unslothai/llama.cpp releases
-# (tag b11030-mix-5ff778e or newer), extracted under
-# /home/dank/GitPackages/unsloth-llama-b11030/. It bundles its own ROCm
-# runtime, so there is no host SDK to gate on and no venv to symlink.
+# Runtime: built from source (scripts/build-kairic-edge-gfx1151.sh), output at
+# /home/dank/GitPackages/ROCmFPX-kairic/build-kairic/bin/llama-server. The
+# Kairic Edge path is a llama.cpp build with the PromptForge sidecar env
+# block (the certified runner's exact env + flags).
 #
-# Shape: one stream, 200k context (operator's call, 2026-09-21), all layers
-# on the iGPU (-ngl 999). The 87 GiB model lives in the iGPU's GTT aperture
-# over the shared LPDDR (gtt_total ~114 GiB), the same regime the CIRU
-# Orca predecessor ran in. The MTP shared-Q8_0 head speculates (1.3-1.7x at
-# concurrency 1, per the model card).
+# Shape: one stream, full 262144 context (the model's native window), all
+# layers on the iGPU (-ngl 999). The 27B dense model (15.5 GiB GGUF + 7.99
+# GiB FFN sidecar + 1.88 GiB GDN sidecar + 0.70 GiB GDN-Output sidecar)
+# fits the iGPU's GTT aperture (~114 GiB).
+#
+# Spec method: native MTP depth 4 (the model's MTP layers) + n-gram drafting
+# (24/64/64). DFlash is NOT available for this model (no DRAFT_DFLASH spec
+# type in the Kairic Edge runtime).
+#
+# Compatibility mode: the harness needs sampling, tools and a reasoning
+# budget, which the fast-greedy path rejects. COMPATIBILITY_MODE=1 enables
+# them (9.7% slower per the build guide). Reasoning is enabled (auto +
+# deepseek format, so thoughts land in reasoning_content).
+#
+# The Qwen3.8 family template keeps prior thinking blocks across turns when
+# preserve_thinking is set; pass --jinja and the kwarg so multi-turn agents
+# do not lose the trail. A request-level chat_template_kwargs still wins.
+# The template gates reasoning_effort to {xhigh, medium, low} — the harness
+# folds its levels onto that set in models.yml (reasoningEffortMap).
 #
 # Tunables come from frank-large.env unless FRANK_LARGE_ENV_FILE overrides
 # it; anything not set there falls through to the defaults below.
@@ -24,85 +37,95 @@ set -euo pipefail
 SCRIPT_DIR="$(cd -- "$(dirname -- "${BASH_SOURCE[0]}")" && pwd)"
 ENV_FILE="${FRANK_LARGE_ENV_FILE:-$SCRIPT_DIR/frank-large.env}"
 if [[ -f "$ENV_FILE" ]]; then
-    set -a
     # shellcheck disable=SC1090
     source "$ENV_FILE"
-    set +a
 fi
 
 # --- defaults (used when the env file does not set them) ------------------
-: "${RUNTIME_DIR:=/home/dank/GitPackages/unsloth-llama-b11030}"
-: "${MODEL_DIR:=/home/dank/models/Qwen3.8-Flash-Next-UD-IQ4_XS}"
+: "${RUNTIME_DIR:=/home/dank/GitPackages/ROCmFPX-kairic/build-kairic}"
+: "${MODEL_DIR:=/home/dank/models/Qwen3.8-27B-IU4-Kairic-Edge}"
 : "${HOST:=0.0.0.0}"
 : "${PORT:=7770}"
-: "${CONTEXT_SIZE:=200000}"
+: "${CONTEXT_SIZE:=262144}"
 : "${PARALLEL_SLOTS:=1}"
-: "${ENABLE_MTP:=1}"
-: "${SPEC_DRAFT_N_MAX:=2}"
-: "${ENABLE_VISION:=0}"
+: "${SPEC_DRAFT_N_MAX:=4}"
+: "${COMPATIBILITY_MODE:=1}"
+: "${CACHE_RAM:=8192}"
 
-SERVER_BIN="$RUNTIME_DIR/llama-server"
+SERVER_BIN="$RUNTIME_DIR/bin/llama-server"
+MODEL_FILE="$MODEL_DIR/Qwen3.8-27B-IU4-Kairic-Edge.gguf"
+FFN_SIDECAR="$MODEL_DIR/Qwen3.8-27B-Kairic-IU4-FFN.pfs"
+GDN_SIDECAR="$MODEL_DIR/Qwen3.8-27B-Kairic-IU4-GDN.pfs"
+GDN_OUTPUT_SIDECAR="$MODEL_DIR/Qwen3.8-27B-Kairic-IU4-GDN-Output.pfs"
 
 if [[ ! -x "$SERVER_BIN" ]]; then
-    echo "missing runtime: $SERVER_BIN" >&2
-    echo "fetch and extract it:" >&2
-    echo "  curl -fL -o /tmp/unsloth-llama.tar.gz https://github.com/unslothai/llama.cpp/releases/download/b11030-mix-5ff778e/app-b11030-mix-5ff778e-linux-x64-rocm-gfx1151.tar.gz" >&2
-    echo "  mkdir -p $RUNTIME_DIR && tar xzf /tmp/unsloth-llama.tar.gz -C $RUNTIME_DIR" >&2
+    echo "ERROR: Kairic Edge runtime not found at $SERVER_BIN" >&2
+    echo "       Build it with: scripts/build-kairic-edge-gfx1151.sh" >&2
     exit 1
 fi
-
-# Point -m at the FIRST split part, not a concatenated file: the parts
-# carry ggml.file_count=3 metadata in part 1's header, so even a
-# byte-correct concatenation still advertises itself as a 3-way split and
-# the loader rejects a non-split filename ("invalid split file name").
-# With -m on part 1, llama.cpp loads the 3-way split natively (parts 2-3
-# are auto-discovered in the same directory).
-MODEL_FILE="$MODEL_DIR/UD-IQ4_XS/Qwen3.8-Flash-Next-UD-IQ4_XS-00001-of-00003.gguf"
-if [[ ! -r "$MODEL_FILE" ]]; then
-    echo "missing target model: $MODEL_FILE" >&2
-    echo "fetch it: bash $SCRIPT_DIR/download-frank-large-unsloth.sh" >&2
-    exit 1
-fi
-
-# The MTP shared head borrows the token embedding and the output projection
-# from the main model at runtime. It must be passed explicitly with -md:
-# sidecar auto-discovery does not search the mtp/ subfolder, so
-# --spec-type draft-mtp on its own would find nothing and serve at the
-# main model's speed with no error saying why.
-MTP_FILE="$MODEL_DIR/mtp/mtp-Qwen3.8-Flash-Next-shared-Q8_0.gguf"
-if [[ "$ENABLE_MTP" == "1" && ! -r "$MTP_FILE" ]]; then
-    echo "missing MTP head: $MTP_FILE" >&2
-    echo "fetch and assemble it: bash $SCRIPT_DIR/download-frank-large-unsloth.sh" >&2
-    exit 1
-fi
-
-# The Qwen3.8 family template keeps prior thinking blocks across turns when
-# preserve_thinking is set; pass --jinja and the kwarg so multi-turn agents
-# do not lose the trail. A request-level chat_template_kwargs still wins.
-# The template gates reasoning_effort to {xhigh, medium, low} — the harness
-# folds its levels onto that set in models.yml (reasoningEffortMap).
-ARGS=(
-    -m "$MODEL_FILE"
-    -c "$CONTEXT_SIZE"
-    -np "$PARALLEL_SLOTS"
-    -ngl 999
-    --jinja
-    --alias frank-large
-    --host "$HOST"
-    --port "$PORT"
-    --chat-template-kwargs '{"preserve_thinking":true}'
-)
-if [[ "$ENABLE_MTP" == "1" ]]; then
-    ARGS+=(-md "$MTP_FILE" --spec-type draft-mtp --spec-draft-n-max "$SPEC_DRAFT_N_MAX")
-fi
-if [[ "$ENABLE_VISION" == "1" ]]; then
-    MMPROJ_FILE="$MODEL_DIR/vision/mmproj-F16.gguf"
-    if [[ ! -r "$MMPROJ_FILE" ]]; then
-        echo "missing projector: $MMPROJ_FILE" >&2
-        echo "fetch and assemble it: bash $SCRIPT_DIR/download-frank-large-unsloth.sh" >&2
+for f in "$MODEL_FILE" "$FFN_SIDECAR" "$GDN_SIDECAR" "$GDN_OUTPUT_SIDECAR"; do
+    if [[ ! -r "$f" ]]; then
+        echo "ERROR: missing asset: $f" >&2
+        echo "       Fetch it with: scripts/download-frank-large-kairic.sh" >&2
         exit 1
     fi
-    ARGS+=(--mmproj "$MMPROJ_FILE")
+done
+
+# --- PromptForge env block (the Kairic Edge runner's exact env) ----------
+export HIP_VISIBLE_DEVICES=0
+export HSA_OVERRIDE_GFX_VERSION=11.5.1
+export LD_LIBRARY_PATH="/opt/rocm/lib:/opt/rocm/llvm/lib:${LD_LIBRARY_PATH:-}"
+export GGML_CUDA_GRAPH_OPT=0
+# Fast-greedy mode (0) rejects sampling/tools/a reasoning budget; compatibility
+# mode (1) enables them. The harness needs all three, so default to 1.
+if [[ "$COMPATIBILITY_MODE" == "1" ]]; then
+    export LLAMA_TARGET_GREEDY_ARGMAX_FASTPATH=0
+    export PROMPTFORGE_OUTPUT_K8_STRICT_GREEDY=0
+else
+    export LLAMA_TARGET_GREEDY_ARGMAX_FASTPATH=1
+    export PROMPTFORGE_OUTPUT_K8_STRICT_GREEDY=1
 fi
+export LLAMA_MTP_CPU_ARGMAX_FASTPATH=1
+export PROMPTFORGE_TARGET_ONLY=0
+export PROMPTFORGE_MODE=iu4_ffn
+export PROMPTFORGE_IU4_SIDECAR="$FFN_SIDECAR"
+export PROMPTFORGE_SIDECAR="$FFN_SIDECAR"
+export PROMPTFORGE_IU4_HADAMARD=1
+export PROMPTFORGE_IU4_SEGMENTED=0
+export PROMPTFORGE_ENABLE_FFN_KEEPERS=0
+export PROMPTFORGE_FFN_KEEPERS=late6
+export PROMPTFORGE_ENABLE_GDN=1
+export PROMPTFORGE_ENABLE_GDN_W8=0
+export PROMPTFORGE_GDN_SIDECAR_OVERRIDE=1
+export PROMPTFORGE_GDN_SIDECAR="$GDN_SIDECAR"
+export PROMPTFORGE_GDN_IU4_HADAMARD=1
+export PROMPTFORGE_ENABLE_GDN_KEEPERS=0
+export PROMPTFORGE_ENABLE_GDN_OUTPUT=1
+export PROMPTFORGE_GDN_OUTPUT_SIDECAR="$GDN_OUTPUT_SIDECAR"
+export PROMPTFORGE_GDN_OUTPUT_KEEPERS=v3_lateq6
+export PROMPTFORGE_ENABLE_SMALLM_IU4=1
+export PROMPTFORGE_ENABLE_SMALLM_GDN_IU4=1
+export PROMPTFORGE_ENABLE_SMALLM_GDN_OUTPUT_IU4=0
+export PROMPTFORGE_ENABLE_IU4_DECODE_M2_M5=0
+
+# --- the serve command (the Kairic Edge runner's flags, fleet-adapted) ----
+ARGS=(
+    -m "$MODEL_FILE" --alias frank-large
+    --host "$HOST" --port "$PORT" --jinja
+    --chat-template-kwargs '{"preserve_thinking":true}'
+    -dev ROCm0 -ngl 999
+    -c "$CONTEXT_SIZE" -b 2048 -ub 512 -fa on -ctk f16 -ctv f16
+    -t 16 -tb 32 -np "$PARALLEL_SLOTS" -ctxcp 32 --cache-ram "$CACHE_RAM"
+    --cache-prompt --cache-idle-slots --metrics
+    --kairic-edge
+    --spec-type draft-mtp
+    --spec-draft-device ROCm0 --spec-draft-ngl 999
+    --spec-draft-type-k f16 --spec-draft-type-v f16
+    --spec-draft-n-max "$SPEC_DRAFT_N_MAX" --spec-draft-n-min 0
+    --spec-draft-p-min 0.0 --spec-draft-p-split 0.10
+    --spec-draft-backend-sampling
+    --temp 0 --top-p 1.0 --top-k 0 --min-p 0.0
+    --reasoning auto --reasoning-format deepseek --reasoning-budget -1
+)
 
 exec "$SERVER_BIN" "${ARGS[@]}"
diff --git a/tests/frank_large_unit.bats b/tests/frank_large_unit.bats
index 5296930..39682f7 100755
--- a/tests/frank_large_unit.bats
+++ b/tests/frank_large_unit.bats
@@ -1,16 +1,18 @@
 #!/usr/bin/env bats
-# Seam: the frank-large unsloth-llama.cpp root unit on dankclaw.
+# Seam: the frank-large Kairic Edge root unit on dankclaw.
 #
-# The 2026-09-21 cutover moved this service from the CIRU v4.4.1 runtime
-# (jcbtc/Qwen3.8-Flash-CIRU-STRIX-Orca) onto the unsloth llama.cpp prebuilt
-# (b11030-mix-5ff778e, ROCm gfx1151) serving
-# unsloth/Qwen3.8-Flash-Next-GGUF:UD-IQ4_XS, still as a ROOT unit: a
-# prebuilt tarball with its own bundled ROCm has no container image to
-# install and no generated unit to dodge. This seam pins what a flag list
-# cannot show: the operator's one-stream / 200k shape, the payload the
-# serve script refuses to start without, the explicit -md MTP wiring
-# (sidecar auto-discovery does not search the mtp/ subfolder), a deploy
-# handler that gates on the port, and the fact that the sweep for a stray
+# The 2026-09-21 cutover moved this service from the unsloth llama.cpp
+# prebuilt (b11030-mix-5ff778e, ROCm gfx1151) serving
+# unsloth/Qwen3.8-Flash-Next-GGUF:UD-IQ4_XS onto the Kairic Edge runtime
+# (ciru-ai/ROCmFPX, branch kairic-edge-qwen38-27b-v1.2, commit 205a3e5)
+# serving jcbtc/Qwen3.8-27B-IU4-Kairic-Edge, still as a ROOT unit: a
+# source-built runtime has no container image to install and no generated
+# unit to dodge. This seam pins what a flag list cannot show: the
+# operator's one-stream / full-window shape, the payload the serve script
+# refuses to start without, the MTP + n-gram spec wiring (DFlash is NOT
+# available for this model), compatibility mode (the harness needs
+# sampling/tools/reasoning, which fast-greedy rejects), a deploy handler
+# that gates on the port, and the fact that the sweep for a stray
 # benchmark run cannot take frank-small down with it.
 
 load helpers
@@ -18,13 +20,15 @@ load helpers
 ENV_FILE="$REPO_ROOT/scripts/frank-large.env"
 UNIT="$REPO_ROOT/scripts/frank-large.service"
 SERVE="$REPO_ROOT/scripts/serve-frank-large.sh"
-FETCH="$REPO_ROOT/scripts/download-frank-large-unsloth.sh"
+FETCH="$REPO_ROOT/scripts/download-frank-large-kairic.sh"
+BUILD="$REPO_ROOT/scripts/build-kairic-edge-gfx1151.sh"
 REMOTE_DEPLOY="$REPO_ROOT/scripts/deploy-remote.sh"
 
-@test "the frank-large unit, env and serve files exist where the deployer installs them from" {
+@test "the frank-large unit, env, serve and build files exist where the deployer installs them from" {
     [ -f "$ENV_FILE" ]
     [ -f "$UNIT" ]
     [ -x "$SERVE" ]
+    [ -f "$BUILD" ]
 }
 
 @test "frank-large is boot-persistent, self-healing and unprivileged" {
@@ -40,61 +44,79 @@ REMOTE_DEPLOY="$REPO_ROOT/scripts/deploy-remote.sh"
     grep -q 'FRANK_LARGE_ENV_FILE=/home/dank/serverAppConfigs/scripts/frank-large.env' "$UNIT"
 }
 
-@test "the operator's shape is one stream of 200k context with MTP on" {
+@test "the operator's shape is one stream of the full 262144 window with MTP depth 4" {
     local ctx slots
     ctx="$(sed -nE 's/^CONTEXT_SIZE=([0-9]+)$/\1/p' "$ENV_FILE")"
     slots="$(sed -nE 's/^PARALLEL_SLOTS=([0-9]+)$/\1/p' "$ENV_FILE")"
-    # 200000 is the operator's requested serving shape (2026-09-21); the
-    # model's trained window is 262144.
-    [ "$ctx" -eq 200000 ]
+    # 262144 is the model's native window (the operator's call, 2026-09-21).
+    [ "$ctx" -eq 262144 ]
     # One stream.
     [ "$slots" -eq 1 ]
-    grep -q '^ENABLE_MTP=1$' "$ENV_FILE"
+    # MTP depth 4 (the model's MTP layers) + n-gram drafting (24/64/64).
+    grep -q '^SPEC_DRAFT_N_MAX=4$' "$ENV_FILE"
 }
 
-@test "frank-large stays text-only until the projector is enabled" {
-    grep -q '^ENABLE_VISION=0$' "$ENV_FILE"
-    # The fleet's image path is frank-small; the projector is on disk and is
-    # one env flip away, but models.yml declares frank-large text-only.
+@test "the harness needs compatibility mode (sampling, tools, reasoning)" {
+    # The fast-greedy path rejects sampling, penalties, probabilities,
+    # grammar-constrained tool calls, logit bias, LoRA and a reasoning
+    # budget. The harness needs all three, so COMPATIBILITY_MODE=1.
+    grep -q '^COMPATIBILITY_MODE=1$' "$ENV_FILE"
+    # The serve script enables the fastpath off in compatibility mode.
+    grep -q 'LLAMA_TARGET_GREEDY_ARGMAX_FASTPATH=0' "$SERVE"
+}
+
+@test "frank-large stays text-only" {
+    # The Kairic Edge model is text-only (no vision projector). The fleet's
+    # image path is frank-small; models.yml declares frank-large text-only.
+    ! grep -q -- '--mmproj' "$SERVE"
     grep -q -- '--mmproj' "$REPO_ROOT/scripts/serve-frank-small.sh"
 }
 
-@test "the serve script pins the prebuilt runtime and refuses a missing build" {
-    grep -q 'unsloth-llama-b11030' "$SERVE"
+@test "the serve script pins the Kairic Edge runtime and refuses a missing build" {
+    grep -q 'ROCmFPX-kairic' "$SERVE"
     grep -q 'llama-server' "$SERVE"
     # A missing runtime or payload is a deploy error with a pointer, not a
     # restart loop.
-    grep -q 'missing runtime' "$SERVE"
-    grep -q 'missing target model' "$SERVE"
-    grep -q 'missing MTP head' "$SERVE"
+    grep -q 'Kairic Edge runtime not found' "$SERVE"
+    grep -q 'missing asset' "$SERVE"
     # The Qwen3.8 template keeps prior thinking blocks across turns only
     # when the kwarg is set.
     grep -q 'preserve_thinking' "$SERVE"
 }
 
-@test "the serve script wires the MTP head explicitly and speculates" {
-    # Sidecar auto-discovery does not search the mtp/ subfolder, so the head
-    # must be passed with -md or the server runs at the main model's speed
-    # with no error saying why.
-    grep -q -- '-md "$MTP_FILE"' "$SERVE"
+@test "the serve script wires the MTP + n-gram spec method and the sidecars" {
+    # The Kairic Edge model uses native MTP depth 4 + n-gram drafting
+    # (24/64/64). DFlash is NOT available for this model (no DRAFT_DFLASH
+    # spec type in the Kairic Edge runtime).
     grep -q -- '--spec-type draft-mtp' "$SERVE"
-    grep -q 'mtp-Qwen3.8-Flash-Next-shared-Q8_0.gguf' "$SERVE"
+    grep -q -- '--kairic-edge' "$SERVE"
+    grep -q -- '--spec-draft-n-max' "$SERVE"
+    # The three PromptForge sidecars (FFN, GDN, GDN-Output) are wired via
+    # the PROMPTFORGE_* env block.
+    grep -q 'PROMPTFORGE_IU4_SIDECAR' "$SERVE"
+    grep -q 'PROMPTFORGE_GDN_SIDECAR' "$SERVE"
+    grep -q 'PROMPTFORGE_GDN_OUTPUT_SIDECAR' "$SERVE"
 }
 
-@test "the payload is fetched resumably and verified against the live Hub tree" {
-    grep -q 'unsloth/Qwen3.8-Flash-Next-GGUF' "$FETCH"
+@test "the payload is fetched resumably and verified against declared sizes + sha256" {
+    grep -q 'jcbtc/Qwen3.8-27B-IU4-Kairic-Edge' "$FETCH"
     grep -q -- '-fL -C -' "$FETCH"
-    # The LFS oids come from the Hub tree API at verify time, not from a
-    # file that could drift; each part's local sha256 is compared to its
-    # oid and a mismatch fails the run.
-    grep -q 'tree/main' "$FETCH"
+    # Each file's local sha256 is compared to its declared oid and a
+    # mismatch fails the run.
     grep -q 'SHA256 MISMATCH' "$FETCH"
 }
 
+@test "the build script pins the Kairic Edge commit and the CK patch" {
+    grep -q 'fdf4bb7fcc984811cef48ce817d89aac064b984a' "$BUILD"
+    grep -q 'composable-kernel-gfx1151-iu4.patch' "$BUILD"
+    grep -q 'gfx1151' "$BUILD"
+}
+
 @test "the fleet endpoint and alias survive the cutover" {
     grep -q '^HOST=0.0.0.0$' "$ENV_FILE"
     grep -q '^PORT=7770$' "$ENV_FILE"
-    grep -q '^MODEL_DIR=/home/dank/models/Qwen3.8-Flash-Next-UD-IQ4_XS$' "$ENV_FILE"
+    grep -q '^MODEL_DIR=/home/dank/models/Qwen3.8-27B-IU4-Kairic-Edge$' "$ENV_FILE"
+    grep -q -- '--alias frank-large' "$SERVE"
 }
 
 @test "the deploy handler retires the halogen quadlet without deleting it" {
@@ -116,7 +138,7 @@ REMOTE_DEPLOY="$REPO_ROOT/scripts/deploy-remote.sh"
     # frank-small on this host is ALSO a llama-server. The sweep must match
     # the model path, not the binary name, or a frank-large deploy restarts
     # the 3090 endpoint.
-    grep -q "pkill -f 'Qwen3.8-Flash-Next-UD-IQ4_XS'" "$REMOTE_DEPLOY"
+    grep -q "pkill -f 'Qwen3.8-27B-IU4-Kairic-Edge'" "$REMOTE_DEPLOY"
     # ...and only inside the large handler: extract it and check there is no
     # bare binary-name pkill in it.
     local large