merge: origin/main c9724b5 into row/A2-D1-mamba-decode-step, for a main-side red

The eighth merge forward, and this one is a REPAIR rather than hygiene. At
`601b576c6`, the commit this branch previously carried, `check-gate-commands`'s
ratchet suite was RED with four failures -- `test_dropping_{cudagraph_dedup,
mtp_depth,residency_config,serve_recipe_args}_from_the_pin_breaks_it` -- each
reporting `ENG-CUDAGRAPH-BREAK` leaving the gated population alongside the row
the case drops.

Attributed by counter-measurement rather than by reading the diff. A detached
worktree at `601b576c6` runs the suite RED at rc=1; the same suite at
`c9724b5ee` runs `Ran 46 tests` `OK` at rc=0. So it is a transient main-side
red, already repaired upstream, and merging forward is the whole fix. This row
touches no gate-command surface.

`.agents/benchmark-record.md` conflicted again and both appended sections are
kept, main's first. `.agents/issue-index.md` was rebuilt as main's file plus
only #1290, #1311 and #1371, with the prefix property asserted.

FOLLOWING_AGENTS_PROTOCOL

Following-Agents-Protocol: true
AI-Assisted: true
Assisted-by: AGENT:claude-opus-5 [Claude Code]

Author: mudler

.agents/benchmark-record.md

@@ -191,6 +191,7 @@ models we already ship + benchmark. Full seam map + M0–M5 W-plan:
 | 15 | `ROAD-V1-D6` | **llama.cpp device breadth folded into scope (user-directed 2026-08-05):** the 11 ggml backends vLLM has no platform for — cann, musa, opencl, openvino, rpc, webgpu, zdnn, zendnn, hexagon, blas, virtgpu — inventoried as `BACKEND-GGML-*`. **SPIKES FIRST:** no implementation before each row's `.agents/specs/<slug>.md` clears the spike contract, per the standing directive. vLLM stays the mirror source; llama.cpp is the breadth reference. | [backend matrix](backend-matrix.md) | ☐ per-row spike required | `INVENTORIED` | first spike accepted |
 | REL | `ROAD-V1-RELEASE` | KISS downloads per OS+host ABI: one adaptive CPU binary and one fat CUDA binary covering every supported SM; per-SM CUDA artifacts are optional diagnostics; stable channels require matching runtime evidence and build-only paths stay preview | [`ENG-RELEASE-BINARIES`](engine-matrix.md), [`ENG-RELEASE-WINDOWS`](engine-matrix.md) | [release binary matrix](specs/release-binary-matrix.md); [Windows pre-alpha extension](specs/windows-binary-release.md) | `ACTIVE` | v0.0.2 published eight primary archive/checksum/provenance triplets plus two indexes from `7020de93652ca920424a10ac5255b34810dd2f24` in run `31466516224` (26 assets). Windows W14-W16 are implemented for one PR; native hosted gates, merged-SHA ten-tuple dry run, matching-hardware evidence, `v0.0.3-pre.1` publication and 32-asset audit remain pending. W12 remains optional/non-primary |
 | IMG | `ROAD-V1-CONTAINERS` | **Published container images on GHCR, built by GitHub Actions (user-directed 2026-08-08).** The same staged bundle `ROAD-V1-RELEASE` defines, shipped from one package `ghcr.io/mudler/vllm.cpp` with the lane in the tag — `:<version>-cuda` / `-vulkan` / `-cpu`, the moving `:latest-cuda` / `:latest-vulkan` / `:latest-cpu`, and a bare `:latest` aliasing the cpu lane — and `ENTRYPOINT vllm-server`. Lanes `cuda` (one fat image, every supported SM), `vulkan`, `cpu`, plus `rocm` blocked-preview; version tags immutable, every `latest-<lane>` moves. Every lane is a `linux/amd64`+`linux/arm64` multi-arch manifest on native runners, because the project's own gate hardware (GB10, Thor, Orin) is arm64. Metal and MLX are NOT-CONTAINERIZABLE and stay static-binary-only — a recorded boundary, not pending work. Depends on the `ROAD-V1-RELEASE` install/stage tree: the image IS the bundle, so the two lanes must not grow separate layouts. No image, workflow or registry package exists. | [`ENG-RELEASE-CONTAINERS`](engine-matrix.md) | [container-images.md](specs/container-images.md) — spec ACCEPTED and **W1-W5/W7 IMPLEMENTED**: one `docker/Dockerfile` whose builder stages call the existing `scripts/build-*-release.sh`, digest-pinned ubuntu runtime bases with the CUDA runtime libs copied and the driver left to the host, ffmpeg in every lane, gated container matrix + image validator + least-privilege publish workflow | `ACTIVE` | **cpu lane BUILT AND GATED e2e 2026-08-10** (783 MB linux/amd64; `/health` 200, `/version` 200, in-container healthcheck, clean SIGTERM, booted on opt-125m). The boot gate immediately found [#312](https://github.com/mudler/vllm.cpp/issues/312): `vllm-server` ignored SIGTERM as PID 1, so `docker stop` hard-killed it (137) after 30 s — now exit 0 in 0.25 s via a self-pipe handler into the existing `server.stop()`. Two silent build-context bugs fixed on the way: `.dockerignore`'s `**/build*/` also matched FILES and was excluding `scripts/build-*-release.sh`, and the builders lacked `file`/`binutils` so the inherited archive validator failed after a full compile. **REMAINING: W6 only** — nothing is published to GHCR, cuda/vulkan are gated statically but never built here, and both arm64 legs are unbuilt so SBSA-vs-Tegra (Thor `sm_110`, Orin `sm_87`) is untouched |
+| QS | `ROAD-V1-QUICKSTART` | **A new reader can run a model without building the tree or already holding a checkpoint.** Two parts. First, `--model` accepts a HuggingFace repository identifier: `org/repo` mirrors vLLM's snapshot download and `org/repo:Q4_K_M` fetches one GGUF file, a form vLLM does not implement. Today `server_main.cpp:433` stores the flag and `model_loader.cpp` opens it as a path, so nothing in the tree fetches a checkpoint, while `docker/Dockerfile:188-192` already sets `HF_HOME=/cache` and declares the volume for a fetch that does not exist and cannot be worked around, because the runtime stage carries no Python and no `curl`. Second, `docs/QUICKSTART.md` gives one `docker run` line that needs no build and no GPU, the `curl` that answers it, and a model table whose every row was run end to end and records the date and host. `README.md:160-176` currently tells the reader to build the tree and to hold a 27B checkpoint, and it names none of the three published lane images. | [engine matrix](engine-matrix.md) | [hf-model-download.md](specs/hf-model-download.md) — spec committed, W1 and W2 landed: the hub protocol and the cache layout, with the DFlash draft path reaching the shared cache walk. No downloader, no `--model` grammar and no quickstart page yet | `READY` | The end-to-end case: `vllm_server_main` with `--model org/repo` against an in-process fake hub boots and completes a request, and turns red when the resolver call site is deleted. Issues [#1280](https://github.com/mudler/vllm.cpp/issues/1280) and [#1281](https://github.com/mudler/vllm.cpp/issues/1281) |
 
 An area row cannot enter `READY` without a real spike under `specs/`, and cannot
 enter `DONE` without exact code and test/evidence anchors. Closed execution