diff --git a/.agents/skills/bump-engines/SKILL.md b/.agents/skills/bump-engines/SKILL.md
index 9c51f582..4680d23e 100644
--- a/.agents/skills/bump-engines/SKILL.md
+++ b/.agents/skills/bump-engines/SKILL.md
@@ -15,10 +15,11 @@ Each engine's `engines/<engine>/AGENTS.md` owns its pin, release id convention a
 
 4. Sweep for the old release id, version and digest. Keep measurement-provenance comments in model files. A bump that changes no flag never needs backend test edits; if one does, the test asserts a committed declaration, which is a coupling bug to fix. When upstream forces an adapter change (step 2), its tests change with it.
 
-5. Qualify on hardware. Flag audits miss default-behaviour changes, which only the gated tests and real deployments catch, so this step is required:
-   * Run `just backend::test-nvidia` on an amd64 NVIDIA host and on GB10.
-   * Deploy each changed engine's profiles on every platform it supports, each on a host whose memory fits its per-device minimum. A row passes when it reaches Ready, answers two completions, shows draft acceptance above 0 where speculative, and launches with the argv the adapter produces.
-   * Report sizing estimates that read low by more than 5% or 500 MiB, measured with the check of the `model-sizing` skill, and a largest fit that doesn't start or lands far below the engine's real start limit; don't refit sizing in a bump unless upstream changed how the engine holds memory, which needs the user's approval.
+5. Qualify on every platform the engine runs on (root `AGENTS.md` "Platform scope"), macOS included. Flag audits miss default-behaviour changes, which only the hardware tests and real deployments catch, so this step is required:
+   * Run the hardware tests in the engine's `AGENTS.md` "Validate" on a host of each platform.
+   * Deploy each changed engine's profiles on one platform, each on a host whose memory fits its per-device minimum. On the other platforms, deploy only the profiles most exposed to the audited upstream changes (drafts, the engine's memory paths) and any that read low; full sweeps cost hours of hardware and many GB of downloads. A row passes when it reaches Ready, answers two completions, shows draft acceptance above 0 where speculative, and launches with the argv the adapter produces.
+   * Report sizing estimates that read low by more than 5% or 500 MiB, measured with the check of the `model-sizing` skill, and a largest fit that doesn't start or lands far below the engine's real start limit; don't refit sizing in a bump unless upstream changed how the engine holds memory, which needs the user's approval. A model that only overruns its margin gets, with approval, a small offset fitted to the measurements already taken, not a new sizing campaign.
    * Run remote hosts one at a time, with one remote command per row: an ssh agent that prompts for approval rejects concurrent signing.
+   * Check a host's free disk before each new model download. The model cache is shared with other workspaces, so report a nearly full disk to the user instead of pruning it.
 
 6. Fix any `AGENTS.md` step that proved wrong.
diff --git a/backend/core/src/model/engine/llama_cpp/binding.rs b/backend/core/src/model/engine/llama_cpp/binding.rs
index 9785946b..3e5a68b4 100644
--- a/backend/core/src/model/engine/llama_cpp/binding.rs
+++ b/backend/core/src/model/engine/llama_cpp/binding.rs
@@ -56,7 +56,7 @@ fn classify(rejections: &[(u16, &str)], status: u16, body: &[u8]) -> Option<Inva
 /// message format, where `%d` stands for a token count.
 ///
 /// A prompt larger than the micro-batch is a server error, and one larger than a slot's context
-/// is a context-size error (b11723 `tools/server/server-context.cpp`).
+/// is a context-size error (b12359 `tools/server/server-context.cpp`).
 const OVERSIZED_PROMPT_REJECTIONS: [(u16, &str); 2] = [
     (
         500,
@@ -69,7 +69,7 @@ const OVERSIZED_PROMPT_REJECTIONS: [(u16, &str); 2] = [
 ];
 
 /// llama.cpp's refusal of a chat prompt larger than a slot's context: the HTTP status it answers
-/// with and its message format (b11723 `tools/server/server-context.cpp`).
+/// with and its message format (b12359 `tools/server/server-context.cpp`).
 const OVERSIZED_CHAT_REJECTIONS: [(u16, &str); 1] = [(
     400,
     "request (%d tokens) exceeds the available context size (%d tokens), try increasing it",
diff --git a/backend/core/src/model/resolved.rs b/backend/core/src/model/resolved.rs
index 4a0bf57d..cc8fe061 100644
--- a/backend/core/src/model/resolved.rs
+++ b/backend/core/src/model/resolved.rs
@@ -1255,7 +1255,7 @@ mod tests {
                 inputs.hardware_class_id.as_str(),
             ),
             (
-                "llama-cpp-2026-10-19",
+                "llama-cpp-2026-11-13",
                 "metal",
                 dotcore_domain::test_support::nix::STORE_PATH,
                 "apple",
diff --git a/backend/core/src/test_support/nix.rs b/backend/core/src/test_support/nix.rs
index e20c94c5..1271206f 100644
--- a/backend/core/src/test_support/nix.rs
+++ b/backend/core/src/test_support/nix.rs
@@ -25,7 +25,7 @@ pub(crate) fn apple_metal_snapshot(model_id: &str) -> ModelDeploymentSnapshot {
         panic!("Apple Metal fixture requires a local snapshot");
     };
     let engine = local.engine_artifact.as_mut().expect("snapshot engine");
-    engine.engine_release_id = "llama-cpp-2026-10-19".to_owned();
+    engine.engine_release_id = "llama-cpp-2026-11-13".to_owned();
     let mut artifact = EngineArtifact {
         selectors: Vec::new(),
         ..engine_artifact("metal", &locked_flake(), "darwin", "arm64", "apple")
diff --git a/backend/core/tests/gated/apple_hardware_integration.rs b/backend/core/tests/gated/apple_hardware_integration.rs
index d0380f28..06e82b13 100644
--- a/backend/core/tests/gated/apple_hardware_integration.rs
+++ b/backend/core/tests/gated/apple_hardware_integration.rs
@@ -129,11 +129,12 @@ fn hw_apple_gpu_is_observed_with_a_fixed_device_id_across_polls() {
         NodeAcceleratorObservation::from_hardware_snapshot("node-a", &snapshots[0], false),
         NodeAcceleratorObservation::available_at(
             "node-a",
-            NodeAcceleratorFacts::apple(),
+            NodeAcceleratorFacts::apple()
+                .with_device_memory(gpu.device_id.clone(), None, None)
+                .with_device_model(gpu.model.clone())
+                .with_host_memory(host_total, host_available),
             observed_at
         )
-        .with_device_memory(gpu.device_id.clone(), None, None)
-        .with_host_memory(host_total, host_available)
     );
 }
 
diff --git a/backend/fixtures/llama-cpp-nvidia-site/catalog/models/stories15m-split-qualification.yaml b/backend/fixtures/llama-cpp-nvidia-site/catalog/models/stories15m-split-qualification.yaml
index 658bfc6f..a1a5178b 100644
--- a/backend/fixtures/llama-cpp-nvidia-site/catalog/models/stories15m-split-qualification.yaml
+++ b/backend/fixtures/llama-cpp-nvidia-site/catalog/models/stories15m-split-qualification.yaml
@@ -21,7 +21,7 @@ artifacts:
 configurations:
   - configurationId: "nvidia-cuda"
     engineArtifacts:
-      - engineReleaseId: "llama-cpp-linux-2026-10-19"
+      - engineReleaseId: "llama-cpp-linux-2026-11-13"
         engineArtifactId: "cuda13"
     artifactBindings:
       - bindingId: "split-q8"
diff --git a/backend/site/src/node/target_inventory.rs b/backend/site/src/node/target_inventory.rs
index 88614fb2..0f492247 100644
--- a/backend/site/src/node/target_inventory.rs
+++ b/backend/site/src/node/target_inventory.rs
@@ -250,6 +250,17 @@ impl NodeAcceleratorFacts {
         self
     }
 
+    /// Names the product model the vendor reports for the device added last.
+    #[cfg(any(test, feature = "test-support"))]
+    pub fn with_device_model(mut self, model: impl Into<String>) -> Self {
+        self.memory
+            .devices
+            .last_mut()
+            .expect("a device to name")
+            .model = model.into();
+        self
+    }
+
     /// Replaces the host memory facts.
     #[cfg(any(test, feature = "test-support"))]
     pub fn with_host_memory(mut self, total_bytes: u64, available_bytes: u64) -> Self {
diff --git a/catalog/engines/llama-cpp-macos.yaml b/catalog/engines/llama-cpp-macos.yaml
index 6b197891..0e96f028 100644
--- a/catalog/engines/llama-cpp-macos.yaml
+++ b/catalog/engines/llama-cpp-macos.yaml
@@ -2,9 +2,9 @@
 # flake as `engines.llama-cpp.native`. The packaged master catalog rewrites `self` to the locked Dotcore
 # revision with precomputed store paths, and fails unless the version equals the engine package
 # version (checked by `checks.dotcore-catalog`).
-engineReleaseId: "llama-cpp-2026-10-19"
+engineReleaseId: "llama-cpp-2026-11-13"
 engineKind: "ENGINE_KIND_LLAMA_CPP"
-version: "b11723"
+version: "b12359" # v1.0.0
 parameterSchema:
   schemaId: "llama-cpp-v2"
   parameters:
diff --git a/catalog/engines/llama-cpp.yaml b/catalog/engines/llama-cpp.yaml
index 675c64f8..1420bc33 100644
--- a/catalog/engines/llama-cpp.yaml
+++ b/catalog/engines/llama-cpp.yaml
@@ -1,8 +1,8 @@
 # Each artifact names a Docker image derivation package; the package pins the exact image.
 # NVIDIA host integration testing remains an explicit responsibility.
-engineReleaseId: "llama-cpp-linux-2026-10-19"
+engineReleaseId: "llama-cpp-linux-2026-11-13"
 engineKind: "ENGINE_KIND_LLAMA_CPP"
-version: "b11723"
+version: "b12359" # v1.0.0
 # Same schema id as the macOS release, so one model configuration can approve artifacts of both.
 parameterSchema:
   schemaId: "llama-cpp-v2"
diff --git a/catalog/models/clef-flash.yaml b/catalog/models/clef-flash.yaml
index 010f7a02..0106bf1b 100644
--- a/catalog/models/clef-flash.yaml
+++ b/catalog/models/clef-flash.yaml
@@ -30,9 +30,9 @@ configurations:
     # The model card's encoder default; the weights accept up to the model's maximum context.
     maximumContextTokens: 16384
     engineArtifacts:
-      - engineReleaseId: "llama-cpp-linux-2026-10-19"
+      - engineReleaseId: "llama-cpp-linux-2026-11-13"
         engineArtifactId: "cuda13"
-      - engineReleaseId: "llama-cpp-2026-10-19"
+      - engineReleaseId: "llama-cpp-2026-11-13"
         engineArtifactId: "metal"
     artifactBindings:
       - bindingId: "q4-k-m"
diff --git a/catalog/models/gemma-4-12b-it-qat.yaml b/catalog/models/gemma-4-12b-it-qat.yaml
index 4ee1d0a5..71f94c41 100644
--- a/catalog/models/gemma-4-12b-it-qat.yaml
+++ b/catalog/models/gemma-4-12b-it-qat.yaml
@@ -40,8 +40,8 @@ configurations:
   # targets and the Metal binary on macOS native targets; both releases share a parameter schema.
   - configurationId: "llama-cpp"
     engineArtifacts:
-      - { engineReleaseId: "llama-cpp-linux-2026-10-19", engineArtifactId: "cuda13" }
-      - { engineReleaseId: "llama-cpp-2026-10-19", engineArtifactId: "metal" }
+      - { engineReleaseId: "llama-cpp-linux-2026-11-13", engineArtifactId: "cuda13" }
+      - { engineReleaseId: "llama-cpp-2026-11-13", engineArtifactId: "metal" }
     artifactBindings:
       - bindingId: "primary"
         modelArtifactId: "primary"
diff --git a/catalog/models/gemma-4-26b-a4b-it.yaml b/catalog/models/gemma-4-26b-a4b-it.yaml
index 3ee2d130..89c38b3d 100644
--- a/catalog/models/gemma-4-26b-a4b-it.yaml
+++ b/catalog/models/gemma-4-26b-a4b-it.yaml
@@ -106,9 +106,9 @@ configurations:
   # targets and the Metal binary on macOS native targets; both releases share a parameter schema.
   - configurationId: "llama-cpp"
     engineArtifacts:
-      - engineReleaseId: "llama-cpp-linux-2026-10-19"
+      - engineReleaseId: "llama-cpp-linux-2026-11-13"
         engineArtifactId: "cuda13"
-      - engineReleaseId: "llama-cpp-2026-10-19"
+      - engineReleaseId: "llama-cpp-2026-11-13"
         engineArtifactId: "metal"
     artifactBindings:
       - bindingId: "q3-xl"
diff --git a/catalog/models/gemma-4-31b-it.yaml b/catalog/models/gemma-4-31b-it.yaml
index 35d9c222..15106d66 100644
--- a/catalog/models/gemma-4-31b-it.yaml
+++ b/catalog/models/gemma-4-31b-it.yaml
@@ -110,9 +110,9 @@ configurations:
   # targets and the Metal binary on macOS native targets; both releases share a parameter schema.
   - configurationId: "llama-cpp"
     engineArtifacts:
-      - engineReleaseId: "llama-cpp-linux-2026-10-19"
+      - engineReleaseId: "llama-cpp-linux-2026-11-13"
         engineArtifactId: "cuda13"
-      - engineReleaseId: "llama-cpp-2026-10-19"
+      - engineReleaseId: "llama-cpp-2026-11-13"
         engineArtifactId: "metal"
     artifactBindings:
       - bindingId: "q3-xl"
diff --git a/catalog/models/gemma-4-e2b.yaml b/catalog/models/gemma-4-e2b.yaml
index 4c73dfed..9e74e68d 100644
--- a/catalog/models/gemma-4-e2b.yaml
+++ b/catalog/models/gemma-4-e2b.yaml
@@ -48,9 +48,9 @@ configurations:
   # targets and the Metal binary on macOS native targets; both releases share a parameter schema.
   - configurationId: "llama-cpp"
     engineArtifacts:
-      - engineReleaseId: "llama-cpp-linux-2026-10-19"
+      - engineReleaseId: "llama-cpp-linux-2026-11-13"
         engineArtifactId: "cuda13"
-      - engineReleaseId: "llama-cpp-2026-10-19"
+      - engineReleaseId: "llama-cpp-2026-11-13"
         engineArtifactId: "metal"
     artifactBindings:
       - bindingId: "q4-k-m"
diff --git a/catalog/models/gemma-4-e4b-it-qat.yaml b/catalog/models/gemma-4-e4b-it-qat.yaml
index 9f3e8a73..ade598eb 100644
--- a/catalog/models/gemma-4-e4b-it-qat.yaml
+++ b/catalog/models/gemma-4-e4b-it-qat.yaml
@@ -39,8 +39,8 @@ configurations:
   # targets and the Metal binary on macOS native targets; both releases share a parameter schema.
   - configurationId: "llama-cpp"
     engineArtifacts:
-      - { engineReleaseId: "llama-cpp-linux-2026-10-19", engineArtifactId: "cuda13" }
-      - { engineReleaseId: "llama-cpp-2026-10-19", engineArtifactId: "metal" }
+      - { engineReleaseId: "llama-cpp-linux-2026-11-13", engineArtifactId: "cuda13" }
+      - { engineReleaseId: "llama-cpp-2026-11-13", engineArtifactId: "metal" }
     artifactBindings:
       - bindingId: "primary"
         modelArtifactId: "primary"
diff --git a/catalog/models/muse-glimmer-30b.yaml b/catalog/models/muse-glimmer-30b.yaml
index 96fddd9a..394fcac6 100644
--- a/catalog/models/muse-glimmer-30b.yaml
+++ b/catalog/models/muse-glimmer-30b.yaml
@@ -50,9 +50,9 @@ configurations:
   # targets and the Metal binary on macOS native targets; both releases share a parameter schema.
   - configurationId: "llama-cpp"
     engineArtifacts:
-      - engineReleaseId: "llama-cpp-linux-2026-10-19"
+      - engineReleaseId: "llama-cpp-linux-2026-11-13"
         engineArtifactId: "cuda13"
-      - engineReleaseId: "llama-cpp-2026-10-19"
+      - engineReleaseId: "llama-cpp-2026-11-13"
         engineArtifactId: "metal"
     artifactBindings:
       - bindingId: "q3-xl"
diff --git a/catalog/models/nanbeige-4-3-33b.yaml b/catalog/models/nanbeige-4-3-33b.yaml
index f2876c3c..53b39b40 100644
--- a/catalog/models/nanbeige-4-3-33b.yaml
+++ b/catalog/models/nanbeige-4-3-33b.yaml
@@ -57,7 +57,7 @@ artifacts:
 configurations:
   - configurationId: "llama-cpp"
     engineArtifacts:
-      - engineReleaseId: "llama-cpp-linux-2026-10-19"
+      - engineReleaseId: "llama-cpp-linux-2026-11-13"
         engineArtifactId: "cuda13"
     artifactBindings:
       - bindingId: "q3-iq3-s"
diff --git a/catalog/models/nemotron-3-5-lightning-30b-a3b.yaml b/catalog/models/nemotron-3-5-lightning-30b-a3b.yaml
index 148c5092..c99fefd9 100644
--- a/catalog/models/nemotron-3-5-lightning-30b-a3b.yaml
+++ b/catalog/models/nemotron-3-5-lightning-30b-a3b.yaml
@@ -45,9 +45,9 @@ configurations:
   # targets and the Metal binary on macOS native targets; both releases share a parameter schema.
   - configurationId: "llama-cpp"
     engineArtifacts:
-      - engineReleaseId: "llama-cpp-linux-2026-10-19"
+      - engineReleaseId: "llama-cpp-linux-2026-11-13"
         engineArtifactId: "cuda13"
-      - engineReleaseId: "llama-cpp-2026-10-19"
+      - engineReleaseId: "llama-cpp-2026-11-13"
         engineArtifactId: "metal"
     artifactBindings:
       - bindingId: "q3-xl"
diff --git a/catalog/models/ornith-1-5-35b-a3b.yaml b/catalog/models/ornith-1-5-35b-a3b.yaml
index 5330190b..1c21647c 100644
--- a/catalog/models/ornith-1-5-35b-a3b.yaml
+++ b/catalog/models/ornith-1-5-35b-a3b.yaml
@@ -421,9 +421,9 @@ configurations:
   # targets and the Metal binary on macOS native targets; both releases share a parameter schema.
   - configurationId: "llama-cpp"
     engineArtifacts:
-      - engineReleaseId: "llama-cpp-linux-2026-10-19"
+      - engineReleaseId: "llama-cpp-linux-2026-11-13"
         engineArtifactId: "cuda13"
-      - engineReleaseId: "llama-cpp-2026-10-19"
+      - engineReleaseId: "llama-cpp-2026-11-13"
         engineArtifactId: "metal"
     artifactBindings:
       - bindingId: "q4-k-m"
@@ -479,7 +479,9 @@ configurations:
         # The prompt cache shares the unified pool with the model.
         parameters:
           promptCacheOnAccelerator: 1
-          baseOffsetBytes: 0
+          # Measured on a GB10 with llama.cpp b12359 (2026-11-13): the derived base terms read up to
+          # 392 MiB low at the profile defaults (Q4_K_M; Q6_K 267 MiB).
+          baseOffsetBytes: 469762048
         bindingSizing:
           # Derived from qwen3-6 q4-xl (GB10): main-weight difference -1,158,046,240 B.
           - bindingId: "q4-k-m"
@@ -513,8 +515,9 @@ configurations:
         # The prompt cache stays in host memory, outside the GPU's own memory.
         parameters:
           promptCacheOnAccelerator: 0
-          # The qwen3-6 dedicated-memory offset (400 MiB, from its RTX 5090 smoke runs).
-          baseOffsetBytes: 419430400
+          # Measured on an RTX 5090 with llama.cpp b12359 (2026-11-13): the derived base terms read up
+          # to 766 MiB low at the profile defaults (Q4_K_M; Q5_K_M 704 MiB, Q6_K 640 MiB).
+          baseOffsetBytes: 889192448
         # The binding parameters are the nvidia-unified values verbatim, as in qwen3-6.
         bindingSizing:
           - bindingId: "q4-k-m"
diff --git a/catalog/models/qwen3-5-4b.yaml b/catalog/models/qwen3-5-4b.yaml
index 4107dfa1..d9967cb9 100644
--- a/catalog/models/qwen3-5-4b.yaml
+++ b/catalog/models/qwen3-5-4b.yaml
@@ -49,9 +49,9 @@ configurations:
   # targets and the Metal binary on macOS native targets; both releases share a parameter schema.
   - configurationId: "llama-cpp"
     engineArtifacts:
-      - engineReleaseId: "llama-cpp-linux-2026-10-19"
+      - engineReleaseId: "llama-cpp-linux-2026-11-13"
         engineArtifactId: "cuda13"
-      - engineReleaseId: "llama-cpp-2026-10-19"
+      - engineReleaseId: "llama-cpp-2026-11-13"
         engineArtifactId: "metal"
     artifactBindings:
       - bindingId: "q3-xl"
diff --git a/catalog/models/qwen3-5-9b.yaml b/catalog/models/qwen3-5-9b.yaml
index 799c98d7..97e8ca14 100644
--- a/catalog/models/qwen3-5-9b.yaml
+++ b/catalog/models/qwen3-5-9b.yaml
@@ -47,9 +47,9 @@ configurations:
   # targets and the Metal binary on macOS native targets; both releases share a parameter schema.
   - configurationId: "llama-cpp"
     engineArtifacts:
-      - engineReleaseId: "llama-cpp-linux-2026-10-19"
+      - engineReleaseId: "llama-cpp-linux-2026-11-13"
         engineArtifactId: "cuda13"
-      - engineReleaseId: "llama-cpp-2026-10-19"
+      - engineReleaseId: "llama-cpp-2026-11-13"
         engineArtifactId: "metal"
     artifactBindings:
       - bindingId: "q3-xl"
diff --git a/catalog/models/qwen3-6-35b-a3b.yaml b/catalog/models/qwen3-6-35b-a3b.yaml
index 36175952..94a7d257 100644
--- a/catalog/models/qwen3-6-35b-a3b.yaml
+++ b/catalog/models/qwen3-6-35b-a3b.yaml
@@ -113,9 +113,9 @@ configurations:
   # One configuration for every platform: CUDA on Linux NVIDIA, Metal on macOS.
   - configurationId: "llama-cpp"
     engineArtifacts:
-      - engineReleaseId: "llama-cpp-linux-2026-10-19"
+      - engineReleaseId: "llama-cpp-linux-2026-11-13"
         engineArtifactId: "cuda13"
-      - engineReleaseId: "llama-cpp-2026-10-19"
+      - engineReleaseId: "llama-cpp-2026-11-13"
         engineArtifactId: "metal"
     artifactBindings:
       - bindingId: "q3-xl"
diff --git a/catalog/models/qwen3-8-27b.yaml b/catalog/models/qwen3-8-27b.yaml
index fe6f00ab..45bc446f 100644
--- a/catalog/models/qwen3-8-27b.yaml
+++ b/catalog/models/qwen3-8-27b.yaml
@@ -89,9 +89,9 @@ configurations:
   # targets and the Metal binary on macOS native targets; both releases share a parameter schema.
   - configurationId: "llama-cpp"
     engineArtifacts:
-      - engineReleaseId: "llama-cpp-linux-2026-10-19"
+      - engineReleaseId: "llama-cpp-linux-2026-11-13"
         engineArtifactId: "cuda13"
-      - engineReleaseId: "llama-cpp-2026-10-19"
+      - engineReleaseId: "llama-cpp-2026-11-13"
         engineArtifactId: "metal"
     artifactBindings:
       - bindingId: "q3-xl"
diff --git a/engines/llama-cpp/AGENTS.md b/engines/llama-cpp/AGENTS.md
index b463b8d0..8397d3a8 100644
--- a/engines/llama-cpp/AGENTS.md
+++ b/engines/llama-cpp/AGENTS.md
@@ -2,38 +2,28 @@
 
 ## How it is built
 
-* Linux: [`docker-image.lock.json`](docker-image.lock.json) pins the upstream image by digest; the
-  node pulls it on first use or loads its stored image. The release declares the driver floor.
-* macOS: `engines.llama-cpp.native` ([`package.nix`](package.nix)), `aarch64-darwin` only. The
-  release tarball is installed unmodified (rewriting the Mach-O files would break their
-  signatures): tools and dylibs under `libexec/llama-cpp/`, linked into `bin/`; the main program is
-  `llama-server`.
+* Linux: [`docker-image.lock.json`](docker-image.lock.json) pins upstream's CUDA image by digest.
+  The release declares the driver floor.
+* macOS: [`package.nix`](package.nix) installs upstream's macOS arm64 release tarball unmodified;
+  rewriting its Mach-O files would break their signatures.
 * The package builds with the repository flake's nixpkgs, so a root nixpkgs bump can change its
   store path; existing deployments keep the store path they pinned.
 * Nothing here is part of root `just build`, `test` or `check`.
-* Shared parameters: [`engines/AGENTS.md`](../AGENTS.md) says where their flags and rules live.
 
 ## Update
 
-1. Pick the newest tag that has both the Linux image and the macOS asset (the two usually move
-   together). Linux image: `just image-lock ghcr.io/ggml-org/llama.cpp:server-cuda13-<tag>`. Images
-   appear after the release tarballs, so the newest tag can lack one; the recipe then fails.
-2. macOS package: check the release has a `llama-<tag>-bin-macos-arm64.tar.gz` asset (some, such as
-   `b10156`, have none). Set `version` in [`package.nix`](package.nix) to the tag, and `src.hash`
-   to the asset's SHA-256 in SRI form (`sha256-...`): the release asset's `digest`, or
-   `nix store prefetch-file` on its URL; convert hex with `nix hash convert --hash-algo sha256 <hex>`.
-3. Declare new releases with the tag as `version`: `llama-cpp-linux-<date>` in `llama-cpp.yaml` and
-   `llama-cpp-<date>` in `llama-cpp-macos.yaml`, after the build's publish date, and move every
-   model approving the old releases to them. Releases are immutable, and the macOS package `version`
-   must equal its release's `version` (`checks.dotcore-catalog` fails otherwise); keep an older
-   release by pinning it to the `github:dotcoreai/dotcore/<rev>` revision that built it.
+* Target upstream's newest stable release (`v<semver>`), not a nightly. It shares its commit with a
+  nightly build tag (`b<build>`): both artifacts are published under that build, and it is the
+  release `version` (note the semver in a comment).
+* Lock the release's CUDA image by its index digest, never by tag. Images can lag the tarballs, and
+  not every nightly gets one.
+* The macOS package `version` must equal its release's `version`; the catalog check fails
+  otherwise.
+* Releases are immutable: declare new ones with the date convention of `llama-cpp.yaml` and
+  `llama-cpp-macos.yaml`, and move every model approving the old ones.
 
 ## Validate
 
-1. `just smoke` (on a Mac): prints the package store path and `llama-server --version`; the build
-   number must match the tag. No model or GPU needed.
-2. Linux image, on an NVIDIA host: `just image-load`, then from the root
-   `just dev::workload-launch <model> <profile> --device nvidia.com/gpu=0` with a small model
-   approving the new release; the workload answers a request.
-3. `just backend::test-nvidia` on an NVIDIA host (from the root) deploys the engine from the
-   workspace's development source through its loader, as a node does.
+* `just smoke` on a Mac checks the package and the build number it reports.
+* The backend's hardware test recipes on each platform in the root "Platform scope":
+  `test-nvidia` (NVIDIA hosts), `test-gb10` (GB10) and `test-apple` (Mac).
diff --git a/engines/llama-cpp/docker-image.lock.json b/engines/llama-cpp/docker-image.lock.json
index 2cd40fac..5e33f018 100644
--- a/engines/llama-cpp/docker-image.lock.json
+++ b/engines/llama-cpp/docker-image.lock.json
@@ -1,46 +1,46 @@
 {
   "repository": "ghcr.io/ggml-org/llama.cpp",
-  "ref": "sha256:fc2894a3be0d37c6f8e4c95f16e62317fecb2649b9572b0922268ba657ef9ad6",
+  "ref": "sha256:18edb904ace42b3aeaac25e5326a4afc4c789cce232f1159a03fc41847d33cf0",
   "platforms": {
     "linux/amd64": {
-      "digest": "sha256:777879aa924d92925ce3c7e6c89ebad75c0d3f1703fc5fd9ed4c0df35ba932de",
-      "manifest": "{\n  \"schemaVersion\": 2,\n  \"mediaType\": \"application/vnd.oci.image.manifest.v1+json\",\n  \"config\": {\n    \"mediaType\": \"application/vnd.oci.image.config.v1+json\",\n    \"digest\": \"sha256:6130e5de11cb62487f4d6bdf1ef552c11bc26bc27bcace4b5f10a441fa2fe434\",\n    \"size\": 20637\n  },\n  \"layers\": [\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:e51aee9c82ec5dd5ba2add49c45c6d85d460512757e2615b69bcdf9469c7cb58\",\n      \"size\": 29763253\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:adbe3366dd3299879e13fee131167e890bc89079fc84e061c2e89605fa6ed616\",\n      \"size\": 10633770\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:3cdc1652d732d475e066bcfffea71065bd2fc4f8d714d47671decb8c17f8f4dc\",\n      \"size\": 392722\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:a00d856b1f6b8821679444fbae19cf78789600b5dc394f8acd7503384654c824\",\n      \"size\": 163053497\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:4f4fb700ef54461cfa02571ae0db9a0dc1e0cdb5577484a6d75e68dc38e8acc1\",\n      \"size\": 32\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:25d13cec859273ace96b18c72354993f005e3438e8b105f56ad0759fc825dbfa\",\n      \"size\": 187\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:541a3eb87cd3c57a7b35621cc132a2096274503bffe4806c14db97943b681042\",\n      \"size\": 6885\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:982bc9a359c56d8ed2787a9aa4682c704c53754c9a42e056cb5de4074bf7cabf\",\n      \"size\": 1720015258\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:e6dd303a971a9838cf56ccb56ff29e238213099b5d87018646089ac5fd4ba412\",\n      \"size\": 60306\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:d6439aa2ea68a6cecdd153d4d3507e68ca1c88d2865868679343af4b0fc5b8b3\",\n      \"size\": 1684\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:4d69f6a802e9b0ff2cecca1632cc3c563151bda148f1839ecb62b543c8436159\",\n      \"size\": 1523\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:a3c33936de5a8816c73428dfcfad464c67bcf8cc020faaa249df24728fbc1c90\",\n      \"size\": 259311752\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:2439865b9c53bf373eb5e3ff69f1874b21d86bf7b6d605e8bff0ecfe57cfb6cf\",\n      \"size\": 153995607\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:26763c40527679373c15ac148267321a9761f6f38fcfb1da4ce7d5e1c86557b1\",\n      \"size\": 25055\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:4f4fb700ef54461cfa02571ae0db9a0dc1e0cdb5577484a6d75e68dc38e8acc1\",\n      \"size\": 32\n    }\n  ],\n  \"annotations\": {\n    \"org.opencontainers.image.created\": \"2026-10-19T04:33:09Z\",\n    \"org.opencontainers.image.description\": \"LLM inference in C/C++\",\n    \"org.opencontainers.image.revision\": \"31288c9c66e59377f7788d920678a565bb89dd37\",\n    \"org.opencontainers.image.source\": \"https://github.com/ggml-org/llama.cpp\",\n    \"org.opencontainers.image.title\": \"llama.cpp\",\n    \"org.opencontainers.image.url\": \"https://github.com/ggml-org/llama.cpp\",\n    \"org.opencontainers.image.version\": \"b11723\"\n  }\n}",
+      "digest": "sha256:7b5faba6707b50d583545ba8714dac8e906927446b1885ccc489cdcb76068961",
+      "manifest": "{\n  \"schemaVersion\": 2,\n  \"mediaType\": \"application/vnd.oci.image.manifest.v1+json\",\n  \"config\": {\n    \"mediaType\": \"application/vnd.oci.image.config.v1+json\",\n    \"digest\": \"sha256:c0046b2fb6170c898cf952006210bc768cbae6de245b7fb11bfe297334c03898\",\n    \"size\": 19890\n  },\n  \"layers\": [\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:7a48b8f0e5c73d0cc2a85b008978396aa19d253dc1b96fdb0b769ee6dcbeca4a\",\n      \"size\": 29767612\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:18735af53e11d0e5997841b58e85111ceb64a1c7d4ebed43c8050f56c94f720e\",\n      \"size\": 4768002\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:3f7f4c851741dde2fb6ef827447b5fed567fa39ee5d66280ed35c09ea847af5b\",\n      \"size\": 436680\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:51d91ac7747ac9c7bf7a47cc378b2ef0fb2cc81fe7d67af839abcb255d35846f\",\n      \"size\": 165522968\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:4f4fb700ef54461cfa02571ae0db9a0dc1e0cdb5577484a6d75e68dc38e8acc1\",\n      \"size\": 32\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:35a8d77d12e86846049b73aefc1d9ca23eee5acb217dcc8904dc0c9923f6e8f4\",\n      \"size\": 186\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:61913d361b867bde8133062f6d92f1e9aab0e033e9fbf0fff656d484b2639dce\",\n      \"size\": 6886\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:f6f7067bb18425fc4c4a62647a618a690da5109c0102e5db5d55b916c08d60af\",\n      \"size\": 1752041121\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:07f6aab61c0d4448e832e1b2ad70444f11d0be721e0121f79b53aad132e9f11e\",\n      \"size\": 60322\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:fdaed1e979bd98b0b5169bb488ae15a3f3b72ce28a425c71483a5544a9b87156\",\n      \"size\": 1683\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:abc658c17cdcf51eb7f4d85912eac6475984fcdb3938f6ac37758139a2deb7aa\",\n      \"size\": 1522\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:480e2cf16e521d7561b77bc89aee9cff4d15f7df857a5ae399822e247955b50c\",\n      \"size\": 258653644\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:d4885731ab4b6c5a81b2e8536d06933f4d3523e6ad88968dae4479ca078ed0a4\",\n      \"size\": 174183641\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:2c2a419e3e91681637f4b9ed88163134d569bfd2bc69984dc899d63f23a57e68\",\n      \"size\": 25167\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:4f4fb700ef54461cfa02571ae0db9a0dc1e0cdb5577484a6d75e68dc38e8acc1\",\n      \"size\": 32\n    }\n  ],\n  \"annotations\": {\n    \"org.opencontainers.image.created\": \"2026-11-13T10:24:49Z\",\n    \"org.opencontainers.image.description\": \"LLM inference in C/C++\",\n    \"org.opencontainers.image.revision\": \"b1b770d90d94cad9e9ea5b993bdbc383a522fb28\",\n    \"org.opencontainers.image.source\": \"https://github.com/ggml-org/llama.cpp\",\n    \"org.opencontainers.image.title\": \"llama.cpp\",\n    \"org.opencontainers.image.url\": \"https://github.com/ggml-org/llama.cpp\",\n    \"org.opencontainers.image.version\": \"v1.0.0-b12359-b1b770d\"\n  }\n}",
       "diffIds": [
-        "sha256:6078cde548a521a729def2ee7875e9f65513c18f0d4bac4db817417617d7006a",
-        "sha256:b743be2fb384a6266c8c355b03f01706b0a146d6464c3be2462fdf155fb0d237",
-        "sha256:d8e11023ba9df4290cb63c12c66934f1ed76d1d4e608c6ac983be971a1d7f338",
-        "sha256:def11fbdaec945f007ca8adb0f009eb1bbf7ca3fba2ef815153803c5a5bc402b",
+        "sha256:bec87777bc2fb2edcdf9af11df1531469e7df911309c1413082b18fcb8610278",
+        "sha256:859495074de220e8c5cfc0705cd6ae6fb530dbe0491da4b6a8f736635baf1118",
+        "sha256:ee791f94b65873771d25f0e206498327bcc57112f8b57fd09c2f1df1bf075921",
+        "sha256:32e5e4c9c46a41e5f73b73577ba34f879f6a80916a53a9ab9df7d43b9746c20c",
         "sha256:5f70bf18a086007016e948b04aed3b82103a36bea41755b6cddfaf10ace3c6ef",
-        "sha256:ffa8b6d26cd31ea21f6f7fa00c932e6d49b6d89dc15a9432d35ab73289b8e14d",
-        "sha256:37916f2ec7845711758cd6254e20712f3145a3aad2d9d8af3901382c4cbf1b9b",
-        "sha256:7ec9cab72963f048a4ec5e8cd34ff025fe483b7824219b2b7042b4a57600e994",
-        "sha256:d2ef768244f7dcbad49a18854263943af397daec80cd68d0d7aa7e3fe32dce97",
-        "sha256:3cc1a6d42a2a2ade7f3c7c07c652f8d20d19940d1e5dca35609e67b2c57b418a",
-        "sha256:4a51ca2cb0836fc3ba8e0aeb9c80b0dfd18eff806e794f184ff28f0e142f32df",
-        "sha256:e3511ac5f7e7f11447eda0d396799f233806453da104c20ed01447acead28092",
-        "sha256:897e7ca26dfa8b6205955c9233f7c5140539f1a19544df047a8f0b656f929a5b",
-        "sha256:15380eae578583caf7db982c8f14dd7fce51f6c56a91f183dab18e8bd3867c35",
+        "sha256:c1d01bd6f2f14fa8245c4ca73cbf4cd87e1a438d449d9718da0a8a633515577e",
+        "sha256:8cd01a34908a39afeb01bed36cf8919bd2eb413d65260277ac89e5a5df70e0ab",
+        "sha256:4fd5732d49e9b7346081b7f2bcef2840f9eca1378154f6b25fce26227164a6f5",
+        "sha256:73f09a9c6ceb1aa0f749a7855bbcd81f43966bddbb2df59d20aca3ddefd507d2",
+        "sha256:9b20092a9029c2363d8ae1537a9f250389d7e350326a78e7494d60d1e7b5f8d9",
+        "sha256:154c9cd64034f90633b9ff02895f1b5308036b94ee4c27bfb7a7818b4ff915d6",
+        "sha256:10950dc6413e17ee61a9dfda660c02f054dfa021860f1c90018e0bbd4e4e40ab",
+        "sha256:236307e704718d104f81a84939ab7d989dd54604120c34927e1dedb07626e45d",
+        "sha256:788e729301e48470bc44e41dc44d5f16aa914137552db7150a1241d2a0e1d89b",
         "sha256:5f70bf18a086007016e948b04aed3b82103a36bea41755b6cddfaf10ace3c6ef"
       ]
     },
     "linux/arm64": {
-      "digest": "sha256:25ffb6197deb81c41f323d8a9e6c6f89beaf58c075ace50162e94a1b5b3265f7",
-      "manifest": "{\n  \"schemaVersion\": 2,\n  \"mediaType\": \"application/vnd.oci.image.manifest.v1+json\",\n  \"config\": {\n    \"mediaType\": \"application/vnd.oci.image.config.v1+json\",\n    \"digest\": \"sha256:e0f2540fe76fcf8313c8f2f65282005749420262c031b26714376d77a78f40ba\",\n    \"size\": 20624\n  },\n  \"layers\": [\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:9332cde1e07015fd0f14e769fc74ea901a957e3f62a76de66638f99854867dd2\",\n      \"size\": 28940015\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:41539e8f3ab4d59d1cfbaf7d1a7b8980174025fddbb4d4d024bde423fcf2d8bd\",\n      \"size\": 10380927\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:2b65f41e87808cf0182588cd6c34f19e0b6a26ffe72c37f66c5ebc9b5d12f886\",\n      \"size\": 407683\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:e9565537a4394232947398643dee33cfa46093d9c4d449c6d4ecc6d89efcfcc3\",\n      \"size\": 160443201\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:cf2558de4d5cf175a4e3006dd6ca332bd29780d432770bb5fc2a0fd09b83b5c2\",\n      \"size\": 160349983\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:d7bed4e05d98a671f84a9e0df60967420a7764e777680a6446a07dd8e107d33e\",\n      \"size\": 187\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:541a3eb87cd3c57a7b35621cc132a2096274503bffe4806c14db97943b681042\",\n      \"size\": 6885\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:85eb39e5062f4c8a710d21d512e64814adbc3fbb2f111a07916e8412b76db5aa\",\n      \"size\": 1782978924\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:d78062fdcaac1deb44e98bc4257ecd327f9c412bbc1d13a843fdbaaaffb5b25b\",\n      \"size\": 60523\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:9c6af9057ef18771ba8a748ace3d875e696b4acb7bf08dba56aec48f51759363\",\n      \"size\": 1683\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:7307e318c047dcb1217ae35330f7489d2c5f4ba6b69f7d4235dd5c3f35155b9a\",\n      \"size\": 1523\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:620d41139e72a3f15c66aae942de5d9cb6648cd78170a53e094761433bf1dc87\",\n      \"size\": 251729076\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:13f5ec667131c0181264cf7d87c38462e3c253772dcac1d725e57924d122f390\",\n      \"size\": 149890005\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:a0aa51d853391f0ce0b2017d8cfc4a5e6c086f69b9b484e38c6b5535b0adc327\",\n      \"size\": 27463\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:4f4fb700ef54461cfa02571ae0db9a0dc1e0cdb5577484a6d75e68dc38e8acc1\",\n      \"size\": 32\n    }\n  ],\n  \"annotations\": {\n    \"org.opencontainers.image.created\": \"2026-10-19T04:33:11Z\",\n    \"org.opencontainers.image.description\": \"LLM inference in C/C++\",\n    \"org.opencontainers.image.revision\": \"31288c9c66e59377f7788d920678a565bb89dd37\",\n    \"org.opencontainers.image.source\": \"https://github.com/ggml-org/llama.cpp\",\n    \"org.opencontainers.image.title\": \"llama.cpp\",\n    \"org.opencontainers.image.url\": \"https://github.com/ggml-org/llama.cpp\",\n    \"org.opencontainers.image.version\": \"b11723\"\n  }\n}",
+      "digest": "sha256:53b4e4ec78a9e78cc5d26b3a90607b7d632885ac3af05243750b6be8bc6456fd",
+      "manifest": "{\n  \"schemaVersion\": 2,\n  \"mediaType\": \"application/vnd.oci.image.manifest.v1+json\",\n  \"config\": {\n    \"mediaType\": \"application/vnd.oci.image.config.v1+json\",\n    \"digest\": \"sha256:962373158ad251b63a3f7bc092bae374971a512bf359df9149efbc499f95581b\",\n    \"size\": 19887\n  },\n  \"layers\": [\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:e2d2bb84d02254f4be480ca3c62f94fc12ac959c11b3ce5bd7da3f411b2b9607\",\n      \"size\": 28940126\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:0f634fbcf9fc8c2356a5cccf53ef14afcf1d5f33d90c643e44bfaf8c0267eac2\",\n      \"size\": 4734603\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:6a23986f6adf4c0b566732ffc47af087df728be8b05f374858c2771117b1ccfc\",\n      \"size\": 450170\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:f669bcd2e8a3c50d0263dfea4c830f7c1ccc3061bbe7351f97660e732f277ea4\",\n      \"size\": 162498706\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:c20fa6cf18399db7371bb19c741acbe434d611528deec3fdc439495c2dd4f486\",\n      \"size\": 159706841\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:f67c82924f2f8b87b3c84e850d6b0cd6ab532e9fccae9e0e4d711f6ff822a3bc\",\n      \"size\": 185\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:757e194b9047b4cf8db7a87070295981994e8082b2248bed1162db9b78886f06\",\n      \"size\": 6886\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:1bb577611d0ae998e5764ce2b710d222c3e6e9dac5eb314e8a9dbe170449e86f\",\n      \"size\": 1928110740\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:78867f5f92e8c1fb45dc47e9785cb90b35408fa17d502141b4276084cf28cf8d\",\n      \"size\": 60546\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:aba5851ef0f3ce1d51a0694348834722df3f385a712487c2f922305a60ef70a0\",\n      \"size\": 1683\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:e2c6749107f0327a758dcab20f2d80d7f506c81d9a94fdd391fc1bfb2b625458\",\n      \"size\": 1522\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:290a9dca27eda624e797065a6a63743420ed090e8ffc71f03cdae4282f74ebd6\",\n      \"size\": 251052282\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:3d37b5b318b4e661b5652bef0bdd3d93a1a93db7051210c2f81f95daed4a32c1\",\n      \"size\": 169915588\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:180052292de254ab2e032e1240ccfcad3af7ecc8ad2ad03e582ab9b431a89d77\",\n      \"size\": 27550\n    },\n    {\n      \"mediaType\": \"application/vnd.oci.image.layer.v1.tar+gzip\",\n      \"digest\": \"sha256:4f4fb700ef54461cfa02571ae0db9a0dc1e0cdb5577484a6d75e68dc38e8acc1\",\n      \"size\": 32\n    }\n  ],\n  \"annotations\": {\n    \"org.opencontainers.image.created\": \"2026-11-13T10:24:53Z\",\n    \"org.opencontainers.image.description\": \"LLM inference in C/C++\",\n    \"org.opencontainers.image.revision\": \"b1b770d90d94cad9e9ea5b993bdbc383a522fb28\",\n    \"org.opencontainers.image.source\": \"https://github.com/ggml-org/llama.cpp\",\n    \"org.opencontainers.image.title\": \"llama.cpp\",\n    \"org.opencontainers.image.url\": \"https://github.com/ggml-org/llama.cpp\",\n    \"org.opencontainers.image.version\": \"v1.0.0-b12359-b1b770d\"\n  }\n}",
       "diffIds": [
-        "sha256:01c818468a75d080ea22eed90a9fa08caeb68931c680c3023b01a0b0427a8641",
-        "sha256:a6259aa424d23c4e5db1e39e9a15735eaa603600920c1474f2954fce5680893f",
-        "sha256:2dc195971afd59315226511925439187e1aad7cac1e0c65ad104282690d0d1c4",
-        "sha256:e9b7a1e2c805195aaa935ded43d4030243f0bb7630a418d1a9717797c371614b",
-        "sha256:ce0ae13b3b07d2692ce3fb76dc24020d775bfa327d7cc19de94aec0497489bbd",
-        "sha256:ddd5ad9424c520444433410191ca0ca1288a3da8a325662e3fc4fe66b1854e91",
-        "sha256:37916f2ec7845711758cd6254e20712f3145a3aad2d9d8af3901382c4cbf1b9b",
-        "sha256:c90abf63ff225c95a9a1057c3d6f931d86dac238f950b65662e50e34f072158a",
-        "sha256:ae9817adee8baad71f2c3419c64001c93564acc35727849461c2c17e82010a55",
-        "sha256:0bd918847131f9f640649a9b0ee5cfdd2b938c261a22a6f4c728a7930d6ae39e",
-        "sha256:54dd08339bcc27d79517960354183b735a6936a012bb632d5351f7c4d156c60f",
-        "sha256:8a8353f16607af0554b20b2f181572f030fd9354043efced609b562ebceeb63f",
-        "sha256:5086d74915f13a60d3353f3e210e316fc1d1a92ae1b4bb417d53065a87f47ad0",
-        "sha256:cd5cfccf07e8994ffa738714d5f4392635ec59e355c4d79d2fe23ef60202d868",
+        "sha256:107ffca6dcd3fc87a7bed8c062f5ddc0d6a4b107205c82c0b49efdb7c1aff967",
+        "sha256:dc6f595091c9c22913165bf1daf78a35e9fde3e1762a147a0e4f199959db0233",
+        "sha256:b1adddf2fb22347fac375c7367cba5dca83cf9b77366b391b5f5fad2cdf69b87",
+        "sha256:fdd12d55290a0b5e3b805a372d335f0842e4062dd27548c79317dfa596701674",
+        "sha256:b8b35307df030d59b597262b04971a7a35eaf84f0f49f004dafaf2c9e63f8fb8",
+        "sha256:39bcf01c92e9c131c7b2c2c8aa41abd44897d57d70f66411dd706c7fb68a0b5c",
+        "sha256:ff6d3af3fb481b59af492fc508077a38b158491ade238d37e1e30a541852be9a",
+        "sha256:8fc57429a7a8f47fd30ed493624f3381225832aa4e44c5fc0ca03a7c4db1fdaa",
+        "sha256:2e7c06d7e8a3eaa7b4606d279822c1096493eb539481c48e38c0b653981e8df4",
+        "sha256:43313c582e418c0899e50e59cdd3a74b69809fc02a149985d0d35a3c9908f059",
+        "sha256:cf1a9323c924334f3b526e1a28762bf7b18c1a5387d18443eca1e93a02bafbfb",
+        "sha256:f0e6271495cfd2216b54d3c182e4d42d718eacec03bc2775a742799482df52f0",
+        "sha256:acdfa6000345162c2bb5a647b75d1a1b619e97de6d0a1051c9c02365caac1f27",
+        "sha256:1379861b46a527ba131ac40110ba859aa1ff7a2dc4d9258f854a019f1c217666",
         "sha256:5f70bf18a086007016e948b04aed3b82103a36bea41755b6cddfaf10ace3c6ef"
       ]
     }
diff --git a/engines/llama-cpp/package.nix b/engines/llama-cpp/package.nix
index 1bf195de..0a2274bb 100644
--- a/engines/llama-cpp/package.nix
+++ b/engines/llama-cpp/package.nix
@@ -8,11 +8,11 @@
 # invalidate their arm64 signatures, so the fixup phase is disabled.
 stdenvNoCC.mkDerivation (finalAttrs: {
   pname = "llama-cpp";
-  version = "b11723";
+  version = "b12359";
 
   src = fetchurl {
     url = "https://github.com/ggml-org/llama.cpp/releases/download/${finalAttrs.version}/llama-${finalAttrs.version}-bin-macos-arm64.tar.gz";
-    hash = "sha256-3rwRu7E8eF1GbqEag303UEdtzVCgZzdYZEaZ91q9hIs=";
+    hash = "sha256-grAKvHQBYOG5mA5sx5wRc0X6NK8FjIDmwKaaqN3M3nU=";
   };
 
   dontConfigure = true;