remote-gpu SKILL.md skillA

remote-gpu from amd-agi/aklr is an agent skill: a SKILL.md folder an AI agent loads when a task calls for it.

Install it, pinned
mdr add amd-agi/aklr/remote-gpu@git:20261117.92345a0

Pins git:20261117.92345a0 in mdr.lock. Pin the fingerprint instead.

Audit grade
A, 17 of 17 checks passed
Kind
SKILL.md skill
Repository
amd-agi/aklr
Path
harness/skills/remote-gpu/SKILL.md
Versions recorded
3
Latest version
git:20261117.92345a0, committed Nov 17, 2026
First version recorded
Nov 5, 2026
Size
10,502 bytes
Repository stars
15
License
MIT

Allocate whole remote GPU nodes on the Slurm-family clusters under ~/.remote-gpu/clusters/ and use them through one `remote-gpu` CLI: get nodes, run commands or docker images, start long background jobs, sync files, release. Use whenever a task needs a real GPU / ROCm machine that isn't local: run/benchmark/profile/test a kernel or model on MI355X/gfx950 or another AMD GPU, multi-node torchrun, "get a node", Slurm, spur, salloc/sbatch/srun, aggressive allocation, or a temporary/preemptible…

Collected from public GitHub. Every version is saved exactly as it was and never changes. Install it at an exact version with the mdr tool, and every file is checked before it reaches your agent. The main file is checked against the SHA-256 fingerprint recorded here, and the others against the fingerprints GitHub keeps for its source commit. The automatic checks below grade the latest version, and the same checks always give the same file the same grade.

What the file says

# remote-gpu — whole GPU nodes on any Slurm-family cluster

Compute nodes can't be reached by ssh, and each scheduler has traps (no GPUs unless asked for,
steps that eat stdin or kill background processes). The CLI handles them, so use it rather than
raw `sbatch`/`srun`/`spur`. It works locally (over ssh) or on a login node:

```bash
RG=${CLAUDE_SKILL_DIR}/scripts/remote-gpu
```

Bash calls don't share variables: write the full path or start each command with `RG=...;`.
Where `${CLAUDE_SKILL_DIR}` is not expanded, it is `harness/skills/remote-gpu` from the AKLR root.

## Arguments

`$ARGUMENTS`:

- Empty or plain language ("2 nodes on mycluster for bench.py"): follow the workflow below.
- A command or `--cluster ...`: run `$RG $ARGUMENTS` once, exactly as given, in the background
  (see **Timeouts**), and once it exits report the `OK`/`ERR` lines and the job handle.

## Clusters

Each cluster is `~/.remote-gpu/clusters/<name>/`, shared by every clone on this machine: `cluster.env`
(login, scheduler, accounts, and what a node is: `GPU_MODEL`, `GPU_ARCH`, `GPUS_PER_NODE`,
`MAX_TIME`, `EXTEND`, `PREEMPTION`) and `cluster.sh` (the scheduler adapter). `$RG help` lists
…

Read the whole file at its exact version.

How to install

No tool to install first
claude plugin marketplace add https://markdownregistry.com/mp/a/art_bulglk3625bliyhj/14aaace678a4ef2b52ff5951d1e85d085b54d1d05e2c88f294efc4a10460f73e.json

Then /plugin install remote-gpu@markdownregistry-bulglk3625bliyhj in Claude Code. It installs git:20261117.92345a0, the one graded A here, pinned to the commit it was graded at, and leaves it there.

The exact contents, by fingerprint
mdr add amd-agi/aklr/remote-gpu@sha256:14aaace678a4ef2b

The line at the top of this page pins the label git:20261117.92345a0, which is what most people want. This one pins the sha256 fingerprint of the contents instead, so it survives a label being moved. Either way the exact version is written to mdr.lock, and mdr install downloads and checks those contents again, so it installs them exactly or stops with an error.

Badge

mdr badge

[![mdr](https://markdownregistry.com/badge/art_bulglk3625bliyhj.svg)](https://markdownregistry.com/a/art_bulglk3625bliyhj)

2 badge image fetches in 30 days. What the badge shows.

Ask from your agent
claude mcp add --transport http markdownregistry https://markdownregistry.com/mcp/registry

Then ask it: what changed in amd-agi/aklr/remote-gpu since the version I have? It reads every saved version here, no key and no account. Any agent that speaks MCP can add it.

Got a copy of remote-gpu from somewhere else? Ask it: where did my remote-gpu come from, and what was changed? It names the public original, the version yours is closest to and every changed line. The text is not kept.

Versions

The registry has recorded 3 versions of this file, between Nov 5, 2026 and Nov 17, 2026, about one every 6 days.

Watch this file (sign in, one email per new version)

versioncommittedcommitsizeaudit
git:20261117.92345a0 latest2026-11-17 92345a0 10,502 BA view · diff
git:20261111.95a87cc2026-11-11 95a87cc 10,128 BA view · diff
git:20261105.1156d712026-11-05 1156d71 9,878 BA view
Keep it current by itself, install once

Committed this skill to a repository? Install the markdownregistry skill updates app on it, or on all of them. When the registry saves a newer version, a pull request brings your copy to it within minutes, the newest version first, then what changed and any check it newly trips. A copy you edited is never touched, and nothing is deleted.

Install on GitHub

It asks to read and write contents and pull requests, and to read metadata. No workflow file, no Actions setting, nothing runs on your runners. Uninstall it and everything the registry kept about your repository is deleted.

Keep it current by itself

Committed this skill to a repository? Add one workflow file, and when a newer version is saved, a pull request brings your copy to it, the newest version first, then what changed and any check it newly trips. A copy you edited is never touched, and nothing is deleted.

mkdir -p .github/workflows && curl -fsSL https://markdownregistry.com/ci/skill-updates.yml -o .github/workflows/skill-updates.yml

It runs daily on GitHub Actions, no key and no account. Read the workflow before you add it.

Check a whole project

From the root of your project, one line hashes each installed skill's SKILL.md, your AGENTS.md and your CLAUDE.md, and asks the registry once. Only the fingerprints and paths are sent.

sha256sum .claude/skills/*/SKILL.md AGENTS.md CLAUDE.md | curl -s --data-binary @- https://markdownregistry.com/api/v1/check-copies

In CI, add --fail-with-body: the answer is 409 when any file is behind, so the step fails and still prints which files.

sha256sum .claude/skills/*/SKILL.md AGENTS.md CLAUDE.md | curl -s --fail-with-body --data-binary @- https://markdownregistry.com/api/v1/check-copies

No sha256sum on your machine? shasum -a 256 prints the same lines.

Audit of the latest version

A  17 of 17 checks passed. Fixed automatic checks with no AI involved, so the same checks give the same answer every run.
  • pass: Frontmatter block present
  • pass: Frontmatter declares a name
  • pass: Frontmatter declares a description
  • pass: Size between 200 bytes and 200 KB (10502 bytes)
  • pass: No zero-width or bidi control characters
  • pass: No instruction hidden inside an HTML comment
  • pass: No link to an exfiltration or paste host
  • pass: No credential-shaped string
  • pass: No instruction to send local credentials anywhere
  • pass: No text hidden with inline styles
  • pass: No prompt-injection phrasing
  • pass: No curl or wget piped into a shell
  • pass: No recursive delete of root, home or parent
  • pass: No instruction to read or print local credentials
  • pass: No base64 blob over 200 characters
  • pass: No link to a raw IP address
  • pass: No script tag

Source

GitHub

amd-agi/aklr · 15 stars · license MIT · pushed 2026-11-19 · branch main

API

GET https://markdownregistry.com/api/v1/artifacts/art_bulglk3625bliyhj
GET https://markdownregistry.com/api/v1/resolve?ref=amd-agi/aklr/remote-gpu
GET https://markdownregistry.com/api/v1/blob/14aaace678a4ef2b52ff5951d1e85d085b54d1d05e2c88f294efc4a10460f73e

Your AI assistant does the legwork. You hear about the deals worth your word. Hand yours the standing instructions at modelranch.com and it joins the network that reads files like this one.

Guides: the SKILL.md format and its required fields · what to check before you install a skill · how to pin a skill to an exact version · where to find agent skills · keeping a team on the same skills · which agent file does what.

More from amd-agi/aklr

amd-aklr Plugin marketplace
amd-agi/aklr · .claude-plugin/marketplace.json · AKLR's agent harness (1 plugin: aklr)
git:20261111.5c4750b · audit B · 15 stars
AGENTS.md AGENTS.md file
amd-agi/aklr · AGENTS.md
git:20261118.5640084 · audit A · 15 stars
optimize-and-deliver-kernels SKILL.md skill
amd-agi/aklr · harness/automations/shepherd/skills/optimize-and-deliver-kernels/SKILL.md · Optimize kernel-less AKLR schemas with KernelForge and deliver them as draft PRs, one round per run (observe, chain…
git:20261117.92345a0 · audit A · 15 stars
setup-shepherd SKILL.md skill
amd-agi/aklr · harness/automations/shepherd/skills/setup-shepherd/SKILL.md · Configure shepherd and prepare its Cursor automation. Use when installing shepherd or changing its repository…
git:20261117.82b2487 · audit A · 15 stars
bad-cases SKILL.md skill
amd-agi/aklr · harness/skills/bad-cases/SKILL.md · A list of mistakes agents have made in AKLR, with what to do instead and how to spot each one in a code change. Use…
git:20261117.82b2487 · audit A · 15 stars
build-runtime-image SKILL.md skill
amd-agi/aklr · harness/skills/build-runtime-image/SKILL.md · Build and validate a reproducible runtime image for an exact AKLR revision and declared ROCm stack. Use when kernel…
git:20261117.82b2487 · audit A · 15 stars
forge-loop-campaign SKILL.md skill
amd-agi/aklr · harness/skills/forge-loop-campaign/SKILL.md · Run an autonomous optimization campaign on a kernel-less AKLR schema with KernelForge on an MI355X (gfx950) machine and…
git:20261117.82b2487 · audit A · 15 stars
geak-campaign SKILL.md skill
amd-agi/aklr · harness/skills/geak-campaign/SKILL.md · Optimize a kernel-less AKLR schema with GEAK (AMD-AGI/GEAK) instead of KernelForge, and deliver the kernel back into…
git:20261117.82b2487 · audit A · 15 stars
kb SKILL.md skill
amd-agi/aklr · harness/skills/kb/SKILL.md · What makes a kernel fast on gfx950 (MI350/MI355X) inside AKLR, and how to write it in FlyDSL, Triton/Gluon or HIP (the…
git:20261117.82b2487 · audit A · 15 stars
kb/flydsl SKILL.md skill
amd-agi/aklr · harness/skills/kb/flydsl/SKILL.md · Writing FlyDSL kernels for AKLR on gfx950 (MI350/MI355X): FlyDSL v0.3.2 limits, compile errors and version differences…
git:20261117.82b2487 · audit A · 15 stars
kb/hip SKILL.md skill
amd-agi/aklr · harness/skills/kb/hip/SKILL.md · Making HIP kernels fast and AITER-exact on gfx950 (MI355X) inside AKLR. Covers memory-bound kernels (normalization…
git:20261117.82b2487 · audit A · 15 stars
kb/triton SKILL.md skill
amd-agi/aklr · harness/skills/kb/triton/SKILL.md · AKLR Triton/Gluon on gfx950/3.8.0: ABI, validation, timing, layouts, memory, MFMA, numerics and operator designs.
git:20261118.ae03491 · audit A · 15 stars

Every file in amd-agi/aklr

Other files named remote-gpu

remote-gpu SKILL.md skill
broomva/skills · skills/compute/remote-gpu/SKILL.md · Orchestrate a headless GPU server (NUC, cloud VM, or any SSH-accessible machine) from a local Mac or workstation. Use…
git:20260628.9edbf07 · audit B · 5 stars

Browse by kind, by grade A, or by owner.