feat(frank-dragoon): open the NPU node the way rootless can actually reach it

The armed boot reached the engine and died in its own preflight:

  halogen npu: /dev/accel/accel0 is here, but this process cannot open it
    (crw-rw---- nobody:nogroup).

The node is root:render 0660 and a rootless container maps exactly one uid/gid,
so `render` is unmapped inside it and `--group-add keep-groups` — the vendor's
advice, and the ComfyUI pattern for /dev/kfd — cannot grant a group the
namespace does not map. Measured 2026-10-19 from a `dank` shell already in
`render`: without the flag denied, with the flag denied, with the udev mode
0666 the open succeeds (OPEN_OK). Same wall 99-comfyui-gpu-access.rules already
solved for the GPU compute nodes, so the accel node gets its counterpart:

  dankclaw/udev/98-halogen-npu-access.rules — KERNEL=="accel[0-9]*",
  SUBSYSTEM=="accel", MODE="0666"

The deployer installs the rule, reloads rules, re-triggers the node and only
then checks the mode; the serve script refuses any node that is not
other-writable, naming the rule, so this cannot return as a crash loop (189
restarts while chat was down is the failure mode being closed).

Both checks read the permission bit directly: `[ -o FILE ]` is not
"other-writable" — it means "exists" and `-o` doubles as the OR operator inside
test, which is why the first cut of the gate refused everything.

Dragoon seam: 33 -> 35 tests. The fixture node is 0666 like the installed one,
a 0660 node is refused with the rule named and nothing rendered, and the rule
file's shape plus the deployer's install/trigger are asserted.

dankclaw/udev/98-halogen-npu-access.rules

@@ -0,0 +1,26 @@
+# /etc/udev/rules.d/98-halogen-npu-access.rules
+#
+# Why the NPU node is world-writable: `frank-dragoon` runs rootless podman under a
+# unit with `User=dank`, and the container's user namespace maps exactly one uid and
+# one gid (dank). `/dev/accel/accel0` is `root:render 0660`, so inside the container
+# the node reads `nobody:nogroup` and the engine dies with its own diagnosis:
+#
+#   halogen npu: /dev/accel/accel0 is here, but this process cannot open it
+#     (crw-rw---- nobody:nogroup).
+#
+# `--group-add keep-groups` is the vendor's advice and it does NOT work in this
+# shape: measured 2026-10-19 from the host as a `dank` shell that was already in
+# `render`, with and without the flag, the node still read `nobody:nogroup` and the
+# open was refused — an unmapped gid cannot be granted to a process inside a
+# namespace that does not map it. Same conclusion `99-comfyui-gpu-access.rules`
+# reached for `/dev/kfd` from the systemd-user-spawn direction: the mode is the only
+# spawn-context-independent door short of running the stack rootful.
+#
+# Blast radius: any process in this host's single uid domain (every rootless
+# container here shares it) can submit to the XDNA2 NPU. The NPU holds no user data
+# and no privileged context — the same trade already accepted for the GPU compute
+# nodes beside it. Installed by `deploy-remote.sh` when the NPU lane is armed and left
+# in place afterwards — the mode costs nothing while the lane is idle. `serve-frank-dragoon.sh`
+# refuses to arm the lane on a node that is not other-accessible, so a host that lost this
+# file fails loudly at boot instead of crash-looping.
+KERNEL=="accel[0-9]*", SUBSYSTEM=="accel", MODE="0666"

scripts/deploy-remote.sh

@@ -1329,6 +1329,9 @@ deploy_anubis() {
 # port gate is the same as every cutover since 2026-09-16: nothing
 # starts until 7770 is released.
 LLM_DRAGOON_IMAGE="${LLM_DRAGOON_IMAGE:-ghcr.io/peonist-ai/halogen-flash-server@sha256:0c61bf84ac22308a53f5d1ca6b86806702d7039e5ebc51cae4c66621b92fe04a}"
+# The payload dir is its own twin because the NPU lane's containment gate below
+# reads it; every artifact path under it stays a full literal as usual.
+LLM_DRAGOON_MODEL_DIR="${LLM_DRAGOON_MODEL_DIR:-/home/dank/models/halogen-v2-abliterated}"
 LLM_DRAGOON_CKPT="${LLM_DRAGOON_CKPT:-/home/dank/models/halogen-v2-abliterated/qwen38-flash-next-v2-abliterated.hgn}"
 LLM_DRAGOON_CKPT_BYTES="${LLM_DRAGOON_CKPT_BYTES:-66687678432}"
 LLM_DRAGOON_NGRAM="${LLM_DRAGOON_NGRAM:-/home/dank/models/halogen-v2-abliterated/qwen38-flash-next-ngram.hgn}"
@@ -1345,7 +1348,7 @@ LLM_DRAGOON_TOK_FILES="${LLM_DRAGOON_TOK_FILES:-tokenizer.json chat_template.jin
 # hang the machine while the fabric clock moves, so halogen refuses to boot with
 # the lane armed unless that clock is held.
 LLM_DRAGOON_NPU_MODELS="${LLM_DRAGOON_NPU_MODELS:-qwen3-embedding-0.6b}"
-LLM_DRAGOON_NPU_DIR="${LLM_DRAGOON_NPU_DIR:-/home/dank/models/halogen-npu}"
+LLM_DRAGOON_NPU_DIR="${LLM_DRAGOON_NPU_DIR:-/home/dank/models/halogen-v2-abliterated/npu}"
 LLM_DRAGOON_DEV_ACCEL="${LLM_DRAGOON_DEV_ACCEL:-/dev/accel/accel0}"
 LLM_DRAGOON_XRT_DIR="${LLM_DRAGOON_XRT_DIR:-/usr/lib/x86_64-linux-gnu}"
 LLM_DRAGOON_NETWORK="${LLM_DRAGOON_NETWORK:-dankclaw_default}"
@@ -1435,12 +1438,34 @@ deploy_llm_dragoon() {
         return 1
     fi
 
-    # --- NPU lane: the same three predicates the serve script checks ---------
+    # --- NPU lane: the same predicates the serve script checks --------------
     if [[ -n "$LLM_DRAGOON_NPU_MODELS" ]]; then
+        # One door into the container: the read-only /models mount of the
+        # payload dir. A staging dir outside it needs a nested bind whose
+        # mountpoint runc cannot create through a read-only parent, so the
+        # parity twin of the serve gate refuses the shape before the restart.
+        if [[ "$LLM_DRAGOON_NPU_DIR/" != "$LLM_DRAGOON_MODEL_DIR"/* ]]; then
+            echo "ERROR: NPU staging dir $LLM_DRAGOON_NPU_DIR is outside the payload dir $LLM_DRAGOON_MODEL_DIR — the lane's files reach the container only through the read-only /models mount" >&2
+            return 1
+        fi
         if [[ ! -e "$LLM_DRAGOON_DEV_ACCEL" ]]; then
             echo "ERROR: NPU node $LLM_DRAGOON_DEV_ACCEL absent — amdxdna is not bound (is iommu=pt still on the boot line?)" >&2
             return 1
         fi
+        # Access, not existence: a rootless container cannot open a root:render
+        # 0660 node (its namespace maps one uid/gid, and keep-groups cannot grant
+        # an unmapped gid — measured 2026-10-19). Install the mode rule, re-trigger
+        # the node, then PROVE the node is other-accessible before the restart.
+        run_privileged install -m 0644 dankclaw/udev/98-halogen-npu-access.rules \
+            /etc/udev/rules.d/98-halogen-npu-access.rules
+        run_privileged udevadm control --reload-rules
+        run_privileged udevadm trigger --sysname-match="$(basename "$LLM_DRAGOON_DEV_ACCEL")"
+        # (`-o FILE` is not "other-writable" — it means "exists", and `-o` is also
+        # the OR operator inside test), so the bit is read straight off the mode.
+        if [[ $(( 8#$(stat -c '%a' "$LLM_DRAGOON_DEV_ACCEL") & 8#2 )) -eq 0 ]]; then
+            echo "ERROR: $LLM_DRAGOON_DEV_ACCEL is still $(stat -c '%a %U:%G' "$LLM_DRAGOON_DEV_ACCEL") after installing the access rule — the lane cannot open it rootless" >&2
+            return 1
+        fi
         for xl in libxrt_coreutil.so.2 libxrt_core.so.2 libxrt_driver_xdna.so.2; do
             if [[ ! -e "$LLM_DRAGOON_XRT_DIR/$xl" ]]; then
                 echo "ERROR: XRT NPU library $LLM_DRAGOON_XRT_DIR/$xl absent (Ubuntu libxrt2 + libxrt-npu2)" >&2

scripts/download-frank-dragoon-npu.sh

@@ -29,7 +29,14 @@ set -euo pipefail
 
 SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
 ENV_FILE="${FRANK_DRAGOON_ENV_FILE:-$SCRIPT_DIR/frank-dragoon.env}"
-DEST="${FRANK_DRAGOON_NPU_DIR:-$HOME/models/halogen-npu}"
+# The env file owns the staging path (it must sit inside the payload dir, which
+# is the container's only /models mount); an exported FRANK_DRAGOON_NPU_DIR wins
+# so the bats seam can stage somewhere disposable.
+DEST="${FRANK_DRAGOON_NPU_DIR:-}"
+if [ -z "$DEST" ] && [ -f "$ENV_FILE" ]; then
+    DEST="$(sed -nE 's/^FRANK_DRAGOON_NPU_DIR=(.*)$/\1/p' "$ENV_FILE" | tail -1)"
+fi
+: "${DEST:?FRANK_DRAGOON_NPU_DIR is not set and $ENV_FILE names no FRANK_DRAGOON_NPU_DIR}"
 HF="https://huggingface.co"
 
 MODELS="${HALOGEN_NPU_MODELS:-}"

scripts/frank-dragoon.env

@@ -87,18 +87,24 @@ HALOGEN_WEIGHTS_LOCK=1
 # fabric-clock dependency, no /v1/embeddings). Rolling the lane back is blanking
 # this one line and re-deploying — nothing else moves.
 # Names are the vendor catalog's folder names (`/opt/halogen/npu/models.txt`
-# inside the image). The engine cannot fetch them here: /models is read-only and
+# inside the image) and the engine reads a model ONLY at /models/npu/<name>/ —
+# HALOGEN_NPU_DIR is gone from the image and setting it is a boot refusal.
+# The engine cannot fetch anything here: /models is read-only and
 # HALOGEN_DOWNLOAD stays unset so the container opens no outbound connection at
-# boot, so scripts/download-frank-dragoon-npu.sh must stage them into
-# $FRANK_DRAGOON_NPU_DIR first (the serve gate and the deployer both refuse by
-# name when a named model is not staged).
+# boot, so scripts/download-frank-dragoon-npu.sh must stage the files first
+# (the serve gate and the deployer both refuse by name when a named model is
+# missing). FRANK_DRAGOON_NPU_DIR therefore lives INSIDE the payload dir: the
+# payload mount is the container's only handle on /models, and runc cannot
+# mkdir a nested /models/npu mountpoint through a read-only bind — the lane's
+# first boot died on precisely that (2026-10-19). Swap the payload dir and this
+# path moves with it.
 # Cost: ~1.5 GiB extra pinned host RAM per named model, and a hard dependency on
 # the host holding the GPU fabric clock — on Strix Halo the iGPU and the NPU
 # together hang the machine (and corrupt NPU results) while FCLK changes speed,
 # so halogen refuses to start with this variable set unless the clock is held
 # (dankclaw/host/halogen-fabric-clock; the deployer installs and gates it).
 HALOGEN_NPU_MODELS=qwen3-embedding-0.6b
-FRANK_DRAGOON_NPU_DIR=/home/dank/models/halogen-npu
+FRANK_DRAGOON_NPU_DIR=/home/dank/models/halogen-v2-abliterated/npu
 # Node paths and the XRT directory are variables so the bats seam can drive
 # these gates on a box with no NPU. Ubuntu's distro XRT (libxrt2 + libxrt-npu2)
 # is mounted FILE BY FILE and TWICE EACH — once where the image looks for XRT,

scripts/serve-frank-dragoon.sh

@@ -125,9 +125,17 @@ fi
 # Empty = the endpoint exactly as it ran before the lane existed: no device, no
 # extra mounts, no gate below. Set = three host preconditions, each refused BY
 # NAME (the 2026-10-01 law: the serve gate and the deployer gate must agree):
-#   1. the staged model files (the engine checks size+sha256 against its own
-#      catalog at boot; it cannot fetch them here because /models is read-only
-#      and HALOGEN_DOWNLOAD stays unset — no outbound connection at boot),
+#   1. the staged model files under $FRANK_DRAGOON_NPU_DIR — a path that MUST
+#      live INSIDE $FRANK_DRAGOON_MODEL_DIR, because the container's only
+#      handle on them is the read-only /models mount. A nested bind at
+#      /models/npu cannot work: runc has to mkdir its mountpoint and the parent
+#      bind is ro (`mkdirat .../merged/models/npu: read-only file system`,
+#      measured 2026-10-19), and the image ships no /models at all for it to
+#      land on. `$MODEL_DIR/npu/<id>` is visible at /models/npu/<id> through
+#      the existing mount, which is also where the vendor's catalog expects it.
+#      The engine re-checks size+sha256 against that catalog at boot; it cannot
+#      fetch anything here because /models is read-only and HALOGEN_DOWNLOAD
+#      stays unset — no outbound connection at boot),
 #   2. the distro XRT's three libraries, each mounted twice (where the image
 #      looks for XRT, and at its own path — the libraries find each other there;
 #      an /opt/xilinx/xrt of symlinks into the system dir dangles inside the
@@ -144,10 +152,31 @@ if [ -n "$NPU_MODELS" ]; then
     NPU_DIR="${FRANK_DRAGOON_NPU_DIR:?HALOGEN_NPU_MODELS is set, so FRANK_DRAGOON_NPU_DIR must name the staged model dir}"
     XRT_DIR="${FRANK_DRAGOON_XRT_DIR:-/usr/lib/x86_64-linux-gnu}"
     SYSFS="${FRANK_DRAGOON_SYSFS:-/sys}"
+    # The lane's files have exactly one door into the container: the read-only
+    # payload mount. A staging dir OUTSIDE it would need a nested bind whose
+    # mountpoint runc cannot create through that read-only parent, so refuse the
+    # shape by name here rather than let runc die at container init (2026-10-19).
+    case "$NPU_DIR/" in
+        "$FRANK_DRAGOON_MODEL_DIR"/*) ;;
+        *) echo "frank-dragoon: FRANK_DRAGOON_NPU_DIR=$NPU_DIR is outside the payload dir $FRANK_DRAGOON_MODEL_DIR — the lane's files reach the container only through the read-only /models mount; move the staging dir inside it" >&2
+           exit 1 ;;
+    esac
     if [ ! -e "$DEV_ACCEL" ]; then
         echo "frank-dragoon: $DEV_ACCEL missing — the NPU lane needs the amdxdna node (is iommu=pt still on the boot line?)" >&2
         exit 1
     fi
+    # Existence is not access. This endpoint runs rootless, so the container's
+    # namespace maps exactly one uid/gid (dank) and a root:render 0660 node reads
+    # `nobody:nogroup` inside it — the engine then dies naming the node itself
+    # (measured 2026-10-19). --group-add keep-groups cannot help: an unmapped gid
+    # cannot be granted into a namespace that does not map it, verified with and
+    # without the flag from a shell already in `render`. The door is the udev mode.
+    # NB: `[ -o FILE ]` is NOT "other-writable" — it means "exists", and `-o` is
+    # also the OR operator inside test — so the permission bit is read directly.
+    if [ $(( 8#$(stat -c '%a' "$DEV_ACCEL") & 8#2 )) -eq 0 ]; then
+        echo "frank-dragoon: $DEV_ACCEL is $(stat -c '%a %U:%G' "$DEV_ACCEL") — a rootless container cannot open it; install dankclaw/udev/98-halogen-npu-access.rules and re-trigger udev (sudo udevadm control --reload-rules && sudo udevadm trigger --sysname-match=accel0)" >&2
+        exit 1
+    fi
     for M in ${NPU_MODELS//,/ }; do
         if [ ! -d "$NPU_DIR/$M" ]; then
             echo "frank-dragoon: NPU model $M not staged in $NPU_DIR;" >&2
@@ -183,7 +212,6 @@ if [ -n "$NPU_MODELS" ]; then
         exit 1
     fi
     NPU_ARGS+=(--device "$DEV_ACCEL" \
-               -v "$NPU_DIR:/models/npu:ro" \
                -e HALOGEN_NPU_MODELS="$NPU_MODELS")
     echo "frank-dragoon: NPU lane armed: $NPU_MODELS on $DEV_ACCEL (fabric clock held)" >&2
 fi

serverAppConfigs-docs/LLM/dankclaw-iommu-translation-off.md

@@ -215,9 +215,10 @@ is in `render`, so a quadlet needs `AddDevice=/dev/accel/accel0` +
 `GroupAdd=keep-groups`. **No NPU endpoint is deployed** —  stays the
 history of the retired endpoint; what changed is only its precondition.
 
-**Deliberately deferred:** the `pp_dpm_fclk` pin (level 7, 2000 MHz) from the issue's
-co-run step. Pinning the fabric clock changes iGPU timing for every workload, and
-nothing co-runs on the NPU yet; it goes in when an NPU endpoint actually exists.
+**The `pp_dpm_fclk` pin (level 7, 2000 MHz) went in on 2026-10-19**, with the first thing that needed it: the
+vendor's root host unit, installed and enabled by `frank-dragoon`'s deployer while its NPU lane is named. Pinning
+the fabric clock changes iGPU timing for every workload, which is exactly why it waited for a co-run to exist; the
+law, the hold and the serve-side re-check live in [[frank-dragoon]].
 PerfOpt still needs Linux 7.4 — `7.0.0-38-generic` has zero `perfopt` symbols and the
 26.04 archive caps at `7.0.0-38.38`.
 

serverAppConfigs-docs/LLM/frank-dragoon.md

@@ -219,7 +219,7 @@ created it — earlier topologies in .
 - Deploy waits for :7770 to be free: a rootless container binds it at CREATE and exits 126, and a
   leftover Quadlet beside the root unit guarantees the race.
 
-## The NPU lane (present in the image, not enabled)
+## The NPU lane (XDNA2, armed by one env var)
 
 XDNA2 executes only vendor-precompiled AIE partition programs — that is the wall the TTS attempt died at
 ([[Frank TTS]]) — and flash-server ships those programs. `HALOGEN_NPU_MODELS=<names>` starts a second engine **in the
@@ -235,11 +235,20 @@ Embedding contract on that lane: `POST /v1/embeddings`, `model: "qwen3-embedding
 (`Instruct: …\nQuery:…`), `base64` accepted, and LangChain's `OpenAIEmbeddings` needs
 `check_embedding_ctx_length=False`.
 
-Four prerequisites the trio does not satisfy today, from the vendor's own host list — and this box is one of the two
-configurations they test (Ubuntu 26.04's XRT 2.21 `libxrt2`+`libxrt-npu2`, the kernel's `amdxdna`, `iommu=pt`):
+Four host prerequisites, each refused **by name** in the serve script and mirrored in `deploy_llm_dragoon` — a
+precondition that surfaces as an engine crash-loop is a gate that was never written. All four are satisfied on this
+box as of 2026-10-19; the endpoint still runs with the lane **off** until that day's mount fix merges, and the first
+armed boot's evidence lands here with it:
 
-1. `--device /dev/accel/accel0` plus `--group-add keep-groups` (the pattern ComfyUI uses for `/dev/kfd`; `dank` is
-   already in `render`).
+1. `--device /dev/accel/accel0` **and** a node the container's namespace can open. `/dev/accel/accel0` is
+   `root:render 0660` and a rootless namespace maps exactly one uid/gid, so the engine's own first words are
+   *"this process cannot open it (crw-rw---- nobody:nogroup)"*. `--group-add keep-groups` — the vendor's advice and
+   the ComfyUI pattern for `/dev/kfd` — is **dead in this shape**: measured 2026-10-19 from a `dank` shell already
+   in `render`, with and without the flag the node stayed `nobody:nogroup` and the open was refused; an unmapped
+   gid cannot be granted across a namespace that does not map it. The door is the udev mode
+   (`dankclaw/udev/98-halogen-npu-access.rules`, installed and re-triggered by the deployer, `0666` like the GPU
+   compute nodes beside it); the serve script refuses any node that is not other-accessible, so the fault cannot
+   return as a crash-loop.
 2. The host's XRT reachable inside: there is **no `/opt/xilinx/xrt`** here — distro packages put the three files the
    container needs (`libxrt_coreutil.so.2`, `libxrt_core.so.2`, `libxrt_driver_xdna.so.2`) in
    `/usr/lib/x86_64-linux-gnu`, where all three exist.
@@ -248,11 +257,18 @@ configurations they test (Ubuntu 26.04's XRT 2.21 `libxrt2`+`libxrt-npu2`, the k
    engine holds the clock itself only when it runs as root with `/sys` mounted; `frank-dragoon` is a **rootless**
    `podman run` under `User=dank`, so the path is the vendor's root host unit (`deploy/host/halogen-fabric-clock`) —
    which is exactly the `pp_dpm_fclk` level-7 pin deferred in [[dankclaw-iommu-translation-off]] with the words "it
-   goes in when an NPU endpoint actually exists". Held: about a watt. Unheld: the hang. Measured 2026-10-19 nothing
-   is held (fabric DPM lists 400–2000 MHz with no forced level, graphics clock sitting at 857 of 2900 MHz).
-4. `HALOGEN_DOWNLOAD=peonist-ai/halogen-qwen3.8-flash-next` on a boot with a **writable** `/models/npu` — the payload
-   mount is `:ro` by law, so the programs land via a throwaway writable run or a dedicated read-write bind for that
-   one subdirectory; the vendor notes a read-only volume works but re-runs the conversion at every start.
+   goes in when an NPU endpoint actually exists". Held: about a watt. Unheld: the hang. The deployer installs and
+   enables that unit whenever the lane is named, and `halogen-fabric-clock status` reports `held (manual)` here as of
+   2026-10-19; the serve script re-reads the same two sysfs files at every start and refuses by name if the top level
+   is not the starred one, because a hold done by hand is a countdown to the next reboot.
+4. The programs are staged under `$FRANK_DRAGOON_NPU_DIR`, which **must sit inside the payload dir**
+   (`/home/dank/models/halogen-v2-abliterated/npu`): the payload's read-only mount is the container's only handle on
+   `/models`. A nested bind at `/models/npu` cannot work — the image ships no `/models`, so runc has to `mkdir` the
+   mountpoint through a read-only parent and dies at container init (measured 2026-10-19: 52 crash-loop restarts with
+   chat down before the mount shape was fixed). `HALOGEN_NPU_DIR` is no escape; the shipped entrypoint rejects it by
+   name ("our NPU models live under `/models/npu/<id>/`"). Nothing downloads at boot — `HALOGEN_DOWNLOAD` stays
+   unset, so the container opens no outbound connection; `scripts/download-frank-dragoon-npu.sh` pins every file by
+   size and sha256 against the catalog and the payload mount carries the tree in.
 
 Price: the programs and their weights sit in host RAM beside the 62.9 GiB of locked weights ("a few GB in all" for
 the default set; ~1.1 GiB for `embeddinggemma-2`), and **the Flash model decodes slower while the NPU works** — same

tests/frank_dragoon_unit.bats

@@ -30,6 +30,7 @@ setup() {
     HOST_UNIT="$ROOT/dankclaw/host/halogen-fabric-clock"
     HOST_SERVICE="$ROOT/dankclaw/host/halogen-fabric-clock.service"
     NPU_FETCH="$ROOT/scripts/download-frank-dragoon-npu.sh"
+    NPU_RULE="$ROOT/dankclaw/udev/98-halogen-npu-access.rules"
     BUILD="$ROOT/scripts/build-rocmfpx.sh"
     TEST_TMP="$(mktemp -d)"
     touch "$TEST_TMP/empty.env"
@@ -357,6 +358,9 @@ $(cat "$TEST_TMP/serve.log")"
     npu_models="$(sed -nE 's/^HALOGEN_NPU_MODELS=(.*)$/\1/p' "$ENV_FILE")"
     grep -qF "LLM_DRAGOON_NPU_DIR=\"\${LLM_DRAGOON_NPU_DIR:-$npu_dir}\"" "$DEPLOYER"
     grep -qF "LLM_DRAGOON_NPU_MODELS=\"\${LLM_DRAGOON_NPU_MODELS:-$npu_models}\"" "$DEPLOYER"
+    # The deployer's own copy of the payload dir is what the lane's containment
+    # gate compares against — a stale twin would wave a bad staging path through.
+    grep -qF "LLM_DRAGOON_MODEL_DIR=\"\${LLM_DRAGOON_MODEL_DIR:-$dir}\"" "$DEPLOYER"
     net="$(sed -nE 's/^FRANK_DRAGOON_NETWORK=(.*)$/\1/p' "$ENV_FILE")"
     grep -qF "LLM_DRAGOON_NETWORK=\"\${LLM_DRAGOON_NETWORK:-$net}\"" "$DEPLOYER"
 }
@@ -415,7 +419,10 @@ $(cat "$TEST_TMP/serve.log")"
 # refused BY NAME instead of surfacing as an engine crash-loop.
 
 npu_fixture() {
+    # 0666 like the installed node: a rootless container's namespace cannot open
+    # the stock root:render 0660 mode, and the serve gate checks exactly that.
     touch "$TEST_TMP/accel0"
+    chmod 666 "$TEST_TMP/accel0"
     mkdir -p "$TEST_TMP/xrt"
     for l in libxrt_coreutil.so.2 libxrt_core.so.2 libxrt_driver_xdna.so.2; do
         touch "$TEST_TMP/xrt/$l"
@@ -469,7 +476,12 @@ $(cat "$TEST_TMP/serve.log")"
     run_with_log
     [ "$status" -eq 0 ]
     [[ "$output" == *"--device $TEST_TMP/accel0"* ]]
-    [[ "$output" == *"-v $TEST_TMP/npu:/models/npu:ro"* ]]
+    # The lane's files ride the payload mount: /models is a read-only bind and
+    # runc cannot mkdir a nested /models/npu mountpoint through it (measured
+    # 2026-10-19), so the staging dir lives INSIDE the payload dir and gets no
+    # -v of its own.
+    [[ "$output" == *"-v $TEST_TMP:/models:ro"* ]]
+    [[ "$output" != *":/models/npu"* ]]
     [[ "$output" == *"-e HALOGEN_NPU_MODELS=qwen3-embedding-0.6b"* ]]
     for l in libxrt_coreutil.so.2 libxrt_core.so.2 libxrt_driver_xdna.so.2; do
         [[ "$output" == *"-v $TEST_TMP/xrt/$l:/opt/xilinx/xrt/lib/$l:ro"* ]]
@@ -479,6 +491,20 @@ $(cat "$TEST_TMP/serve.log")"
     [[ "$output" == *"-e HALOGEN_CHECKPOINT=/models/qwen.hgn"* ]]
 }
 
+@test "the lane refuses a staging dir outside the payload dir and says why" {
+    # A staging dir outside the payload dir needs a nested bind the read-only
+    # /models mount cannot host: runc dies at container init, which is AFTER the
+    # deployer has stopped the old unit — so the shape is refused here, before
+    # anything is rendered.
+    podman_stub
+    good_env
+    npu_fixture
+    run_with_log FRANK_DRAGOON_NPU_DIR=/elsewhere/npu
+    [ "$status" -ne 0 ]
+    [[ "$output" == *"/elsewhere/npu"*"outside the payload dir"* ]]
+    [[ "$output" != *"--device"* ]]
+}
+
 @test "the lane refuses a model that is not staged and names the fetch script" {
     # The engine cannot fetch these at boot here (/models is read-only and
     # HALOGEN_DOWNLOAD stays unset), so an un-staged name must stop the unit
@@ -504,6 +530,32 @@ $(cat "$TEST_TMP/serve.log")"
     [[ "$output" == *"iommu=pt"* ]]
 }
 
+@test "the lane refuses an NPU node the container namespace could not open, and names the rule" {
+    # Stock Debian mode is root:render 0660. The engine's own first words on such
+    # a node are "this process cannot open it (crw-rw---- nobody:nogroup)" and no
+    # flag fixes a mode an unmapped gid cannot cross (keep-groups measured dead
+    # 2026-10-19), so the unit refuses instead of crash-looping 189 times.
+    podman_stub
+    good_env
+    npu_fixture
+    chmod 660 "$TEST_TMP/accel0"
+    run_with_log
+    [ "$status" -ne 0 ]
+    [[ "$output" == *"$TEST_TMP/accel0 is 660"* ]]
+    [[ "$output" == *"98-halogen-npu-access.rules"* ]]
+    [[ "$output" != *"--device"* ]]
+}
+
+@test "the NPU access rule is the accel counterpart of the GPU compute rule" {
+    [ -f "$NPU_RULE" ]
+    grep -qE '^KERNEL=="accel\[0-9\]\*", SUBSYSTEM=="accel", MODE="0666"$' "$NPU_RULE"
+    # The deployer installs the rule and re-triggers the node BEFORE it trusts
+    # the mode, so a machine that never had it self-heals instead of refusing.
+    grep -qE '98-halogen-npu-access\.rules' "$DEPLOYER"
+    grep -qE 'udevadm control --reload-rules' "$DEPLOYER"
+    grep -qE 'udevadm trigger --sysname-match=' "$DEPLOYER"
+}
+
 @test "the lane refuses when an XRT NPU library is missing" {
     podman_stub
     good_env
@@ -560,6 +612,9 @@ $(cat "$TEST_TMP/serve.log")"
     grep -qE 'download-frank-dragoon-npu\.sh' "$DEPLOYER"
     # The lane's own health gate is a real embedding, not a 200 on /health.
     grep -qE '/v1/embeddings' "$DEPLOYER"
+    # Parity with the serve gate: a staging dir outside the payload dir is
+    # refused before the restart, not left to runc.
+    grep -qE 'outside the payload dir' "$DEPLOYER"
 }
 
 @test "the fabric-clock host unit is the vendor shape and boots on its own" {