hyperloom.inference_optimizer.multi_node SKILL.md skillA
hyperloom.inference_optimizer.multi_node from amd-agi/hyperloom is an agent skill: a SKILL.md folder an AI agent loads when a task calls for it.
mdr add amd-agi/hyperloom/hyperloom.inference_optimizer.multi_node@git:20261028.511deb9Pins git:20261028.511deb9 in mdr.lock. Pin the fingerprint instead.
- Audit grade
- A, 17 of 17 checks passed
- Kind
- SKILL.md skill
- Repository
- amd-agi/hyperloom
- Path
src/hyperloom/inference_optimizer/multi_node/SKILL.md- Versions recorded
- 12
- Latest version
- git:20261028.511deb9, committed Oct 28, 2026
- First version recorded
- Jul 28, 2026
- Size
- 17,604 bytes
- Repository stars
- 262
- License
- a license file GitHub does not identify
Multi-node companion to the inference_optimizer skill. Use when the user prompt asks for inference optimization that needs more GPU / memory than a single pod provides (i.e. ``nodes >= 2``) — typical prompt signals are ``Nodes=N`` / ``N pods`` / ``TP=N`` larger than one pod's GPU count, or any model that cannot fit on one pod's GPUs. Drives a platform-provisioned multi-node cluster through the ``hyperloom multi-node`` CLI.
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
# Multi-Node Skill (infera + rayjob backends) **The cluster is not created here.** The platform provisions it (an `InferaDeployment` or a `RayJob`) before the optimizer starts and hands it over through the `HYPERLOOM_MN_EXT_*` env vars (see "Cluster hand-off"). This CLI drives an **already-running** cluster and never creates or tears one down; the platform reclaims it when the session ends. Drive every action through the Python CLI. **Never `ray.init`, `kubectl`, or raw `curl` to a pod.** All state lives in the file resolved from `$MULTI_NODE_STATE_FILE`, else `$INFERENCE_OPTIMIZER_CURRENT_SESSION_DIR/runtime/multi_node_state.json` — one of the two MUST be set or the CLI raises (sandbox-local; synthesized from the hand-off env on first use, rewritten by each subcommand). Re-read it every turn — never cache `service_url` / `head_pod_ip` across actions (a re-provisioned cluster rewrites them). | Action | Use | Never | |----------------|-------------------|-------------------------------------------| | Restart server | `restart-server` | `kubectl exec … sglang.launch_server` | …
Read the whole file at its exact version.
How to install
mdr add amd-agi/hyperloom/hyperloom.inference_optimizer.multi_node@sha256:f2fced77028c23c5The line at the top of this page pins the label git:20261028.511deb9, 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.
[](https://markdownregistry.com/a/art_kegh463fkpfpo2qe)
3 badge image fetches in 30 days. What the badge shows.
Versions
The registry has recorded 12 versions of this file, between Jul 28, 2026 and Oct 28, 2026, about one every 8 days. Its first recorded version is graded B and its latest A.
| version | committed | commit | size | audit | |
|---|---|---|---|---|---|
| git:20261028.511deb9 latest | 2026-10-28 | 511deb9 | 17,604 B | A | view · diff |
| git:20261026.3d39df3 | 2026-10-26 | 3d39df3 | 17,461 B | A | view · diff |
| git:20261019.f7f64ac | 2026-10-19 | f7f64ac | 16,787 B | A | view · diff |
| git:20261018.8f7be91 | 2026-10-18 | 8f7be91 | 16,799 B | A | view · diff |
| git:20261013.6df52ff | 2026-10-13 | 6df52ff | 16,816 B | A | view · diff |
| git:20261009.976d7a0 | 2026-10-09 | 976d7a0 | 16,853 B | A | view · diff |
| git:20260917.0d8dd3b | 2026-09-17 | 0d8dd3b | 16,823 B | A | view · diff |
| git:20260909.f705855 | 2026-09-09 | f705855 | 17,370 B | A | view · diff |
| git:20260819.d10c098 | 2026-08-19 | d10c098 | 17,403 B | A | view · diff |
| git:20260803.be96432 | 2026-08-03 | be96432 | 17,165 B | A | view · diff |
| git:20260802.980063c | 2026-08-02 | 980063c | 16,862 B | A | view · diff |
| git:20260728.dd3b07b | 2026-07-28 | dd3b07b | 34,799 B | B | view |
Audit of the latest version
- pass: Frontmatter block present
- pass: Frontmatter declares a name
- pass: Frontmatter declares a description
- pass: Size between 200 bytes and 200 KB (17604 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
amd-agi/hyperloom · 262 stars · license a license file GitHub does not identify · pushed 2026-11-06 · branch main
API
GET https://markdownregistry.com/api/v1/artifacts/art_kegh463fkpfpo2qe GET https://markdownregistry.com/api/v1/resolve?ref=amd-agi/hyperloom/hyperloom.inference_optimizer.multi_node GET https://markdownregistry.com/api/v1/blob/f2fced77028c23c5b77865d41891d11e5e3a161ba850353d66e8ba17ebb8563f
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.