Add Windows to scope

Author: danielholanda

skills/rocm-doctor/SKILL.md

@@ -1,22 +1,25 @@
 ---
 name: rocm-doctor
 description: >-
-  Diagnoses why ROCm, PyTorch, or llama.cpp fails on an AMD GPU by
-  matching symptoms against a closed catalog of known misconfigurations,
-  then either applies a low-risk fix with consent or hands back the
-  exact next step. Also routes Lemonade, LM Studio, and Ollama users to
-  the right upstream channel. Use when the user says "ROCm/HIP isn't
-  working", "torch.cuda.is_available() is False on Radeon/Ryzen AI",
-  "rocminfo can't find my GPU", "hipErrorNoBinaryForGpu",
+  Diagnoses why ROCm, the HIP SDK, PyTorch, or llama.cpp fails on an AMD
+  GPU by matching symptoms against a closed catalog of known
+  misconfigurations on Linux and Windows, then either applies a low-risk
+  fix with consent or hands back the exact next step. Also routes
+  Lemonade, LM Studio, and Ollama users to the right upstream channel.
+  Use when the user says "ROCm/HIP isn't working", "torch.cuda.is_available()
+  is False on Radeon/Ryzen AI", "rocminfo can't find my GPU",
+  "hipInfo.exe can't see my Radeon", "amdhip64_6.dll could not be found",
+  "vcruntime140_1.dll missing", "HIP SDK installer left things broken",
+  "Adrenalin driver too old for the HIP SDK", "hipErrorNoBinaryForGpu",
   "HSA_STATUS_ERROR_INVALID_ISA", "invalid device function",
   "Unable to open /dev/kfd", "ROCk module is NOT loaded",
   "libamdhip64.so cannot open shared object file", "ROCm wheel doesn't
   see my gfx1151/gfx1150/gfx1103 (Strix Halo, Phoenix)", "iGPU/dGPU
   collision", "multi-GPU hang on AMD"; or mentions HSA_OVERRIDE_GFX_VERSION,
-  HIP_VISIBLE_DEVICES, PYTORCH_ROCM_ARCH, render-group / /dev/kfd
+  HIP_VISIBLE_DEVICES, HIP_PATH, PYTORCH_ROCM_ARCH, render-group / /dev/kfd
   permissions, amdgpu blacklist, Secure Boot, or asks where to file a
   Lemonade / LM Studio / Ollama issue. Do NOT use for non-AMD GPUs,
-  fresh installs, or performance tuning.
+  fresh installs, performance tuning, or ROCm-on-WSL2.
 ---
 
 # ROCm Doctor
@@ -50,38 +53,63 @@ Out of scope:
 
 - NVIDIA / Intel / Apple Silicon GPUs. Exit cleanly and tell the user.
 - Fresh installs on a clean machine. That's a setup task; point at
-  [`amdgpu-install`](https://rocm.docs.amd.com/projects/install-on-linux/en/latest/install/install-overview.html).
+  [`amdgpu-install`](https://rocm.docs.amd.com/projects/install-on-linux/en/latest/install/install-overview.html)
+  (Linux) or the [HIP SDK installer](https://www.amd.com/en/developer/resources/rocm-hub/hip-sdk.html)
+  (Windows).
 - Pure performance complaints. Those belong in `mi-tuner` /
   `omniperf-tune` / `apu-memory-tuner`.
+- **WSL2** (running Linux on top of Windows). The ROCm-on-WSL flow needs
+  Adrenalin Pro plus the WSL kernel update on the Windows host -- those
+  failure modes are not in this catalog. `examine.py` detects WSL via
+  `/proc/version` and exits 2 with a route-out message; if the user wants
+  WSL specifically, point them at <https://rocm.docs.amd.com/projects/install-on-wsl/en/latest/>.
 
 ## Prerequisites
 
-- **OS:** Linux. Phase 0 is Linux-only; on Windows, the HIP SDK / Adrenalin
-  path is its own ecosystem and this skill cannot help.
-- **Tools the agent will invoke as part of examination** (best-effort; the
-  script degrades when one is missing):
+- **OS:** Linux **or** Windows (native). The catalog has 12 Linux entries
+  (5 of which are also valid on Windows) and 3 Windows-only entries; the
+  scripts pick the right subset for the host they run on.
+- **Linux tools the agent will invoke as part of examination** (best-effort;
+  the script degrades when one is missing):
   - `lspci` (always present on desktop distros)
   - `rocminfo` (when ROCm is installed)
   - `journalctl` or `dmesg` (for amdgpu kernel-ring evidence)
   - `python` / `python3` to introspect PyTorch
   - `llama-cli` / `llama-server` / `main` to introspect llama.cpp
+- **Windows tools the agent will invoke as part of examination**:
+  - `powershell` (always present on Windows 10+) for `Get-CimInstance
+    Win32_VideoController` / `Win32_Processor` and the env-scope reads.
+  - `hipInfo.exe` from `%HIP_PATH%\bin` -- the Windows analog of `rocminfo`.
+    Absence is itself a signal (see `fix-13-hip-sdk-missing`).
+  - `setx` for env-var persistence and User-PATH edits (analog of editing
+    `~/.bashrc` on Linux).
+  - `python` to introspect PyTorch.
 - **Permissions:** examination is fully read-only and works as a regular
-  user. Several fixes need `sudo` (the recipe metadata flags this); the
-  script always prints the command before asking for consent.
+  user on both OSes. Linux fixes that need `sudo` are flagged in their
+  recipe metadata; Windows fixes that touch the Machine env scope are
+  flagged similarly and `apply_fix.py` does NOT self-elevate -- the user
+  has to run an Administrator PowerShell when those are required.
 
 Silent footguns to surface explicitly when relevant:
 
 - `HSA_OVERRIDE_GFX_VERSION` -- forcing an unsupported gfx target works
-  for `rocminfo` but causes page faults at runtime. Diagnosis
+  for `rocminfo`/`hipInfo` but causes page faults at runtime. Diagnosis
   `fix-2-unset-override` is the response when this is set on a GPU that
-  already has a native wheel.
+  already has a native wheel; on Windows it can be persisted in either
+  the User or Machine env scope, so check both.
 - `HIP_VISIBLE_DEVICES` -- on dual-GPU systems (APU + dGPU) the iGPU is
   often index 0 and destabilises HIP unless explicitly hidden.
+- `HIP_PATH` (Windows) -- if the user has multiple HIP SDK versions
+  installed under `C:\Program Files\AMD\ROCm\`, `HIP_PATH` decides which
+  one PyTorch / hipInfo actually loads. Pointing it at the wrong major
+  produces the same failure mode as `fix-8-wheel-rocm`.
 - `PYTORCH_ROCM_ARCH` -- only honored during a *build* of PyTorch. Setting
   it at runtime does nothing for a prebuilt wheel.
-- `LD_LIBRARY_PATH` -- a wheel-bundled `libamdhip64.so` shadowed by a
-  system one (or vice versa) gives confusing `cannot open shared object
-  file` errors that look like fix-8 but are really a load-order bug.
+- `LD_LIBRARY_PATH` (Linux) -- a wheel-bundled `libamdhip64.so` shadowed
+  by a system one (or vice versa) gives confusing `cannot open shared
+  object file` errors that look like fix-8 but are really a load-order
+  bug. The Windows analog is `PATH` order: a stale HIP SDK bin directory
+  earlier on PATH than the one matching `HIP_PATH`.
 
 ## The three-step flow
 
@@ -114,7 +142,7 @@ script pick. Exit codes:
 | Exit | Meaning | Next action |
 |---|---|---|
 | 0 | Examined; AMD GPU present | Continue to Step 2. |
-| 2 | Not Linux / no AMD GPU | Stop. Route the user. |
+| 2 | Wrong platform (WSL, neither Linux nor Windows, no AMD GPU) | Stop. Route the user. |
 | 3 | Probes partially failed | Continue but warn the user. |
 
 For a quick read-only summary without piping JSON, drop `--json`:
@@ -123,13 +151,16 @@ For a quick read-only summary without piping JSON, drop `--json`:
 python scripts/examine.py --framework pytorch
 ```
 
-`examine.py` collects exactly the facts the diagnosis catalog needs:
-OS / kernel, AMD GPUs and gfx targets, `amdgpu` / `amdkfd` status,
-`/dev/kfd` ownership and group, user's group membership, system ROCm
-version and install method, framework version and arch list, the
+`examine.py` collects exactly the facts the diagnosis catalog needs.
+On Linux: OS / kernel, AMD GPUs and gfx targets, `amdgpu` / `amdkfd`
+status, `/dev/kfd` ownership and group, user's group membership, system
+ROCm version and install method, framework version and arch list, the
 silent-footgun env vars, container/IOMMU state, and recent `amdgpu`
-kernel log lines. It deliberately does NOT spawn heavy probes (no kernel
-launches, no model downloads).
+kernel log lines. On Windows: AMD adapters and gfx targets via
+`Win32_VideoController` + `hipInfo.exe`, the HIP SDK install path and
+version, the Adrenalin / kernel-mode driver version, MSVC redistributable
+presence, and the same env-var snapshot. It deliberately does NOT spawn
+heavy probes (no kernel launches, no model downloads).
 
 ### Step 2: diagnose
 
@@ -204,16 +235,18 @@ useless probes against the host.
 
 | Framework | Examine the host? | Action |
 |---|---|---|
-| PyTorch | Yes | `python scripts/examine.py --framework pytorch`, then `diagnose.py`. |
-| llama.cpp (built against system ROCm) | Yes | `python scripts/examine.py --framework llama-cpp`, then `diagnose.py`. |
+| PyTorch (Linux ROCm wheel) | Yes | `python scripts/examine.py --framework pytorch`, then `diagnose.py`. |
+| PyTorch (Windows TheRock wheel) | Yes | Same scripts; on Windows the catalog filters to Linux+Windows + Windows-only entries. |
+| llama.cpp (built against system ROCm/HIP SDK) | Yes | `python scripts/examine.py --framework llama-cpp`, then `diagnose.py`. |
 | Lemonade | No -- ships its own ROCm | Route to <https://github.com/lemonade-sdk/lemonade/issues> and the Lemonade [Discord](https://discord.gg/5xXzkMu8Zk). |
 | LM Studio | No -- ships its own runtime | Route to <https://lmstudio.ai/docs/app> (in-app support; no public repo). |
 | Ollama | No -- ships its own runtime | Route to <https://github.com/ollama/ollama/issues> and the Ollama Discord. |
 | vLLM / SGLang | Out of scope until phase 1+ | Route to the project's own issue tracker. |
 
 If a Lemonade / LM Studio / Ollama user *does* have a host-level ROCm
 problem (rare), it shows up when their app fails AND a standalone
-`rocminfo` also fails. Only then escalate to the full examination.
+`rocminfo` (Linux) / `hipInfo.exe` (Windows) also fails. Only then
+escalate to the full examination.
 
 ## Safety rules