agentmods.dev

deploy-llm-d

deploy-llm-d is a skill for Claude Code, Codex from llm-d-incubation/llm-d-skills. It costs 71 tokens per session (6,152 once invoked), scanned A, original, Apache-2.0.

A procedure for configuring and deploying llm-d, a system for serving language models, on an existing Kubernetes or OpenShift cluster. It uses Well-Lit Path deployment guides and limits changes to the chosen namespace.

In plain words
What is it for?
Setting up llm-d, checking or customizing its deployment, and creating resources such as namespaces, storage, routes, or Helm releases when requested.
Why use it?
Deploying a model-serving system involves many cluster resources and can accidentally affect unrelated workloads. This keeps the work scoped and explains commands before running them.

Skill for Claude CodeCodex

Written for no agent in particular: nothing here depends on one.

Needs its repository: it reads a path above its own folder, which exists only inside the repository. The line is - ../../llm-d/guides/<guide>/modelserver/<ACCELERATOR_TYPE>/<MODEL_SERVER>/<INFRA_PROVIDER> # relative to deployments/<name>/; kustomize rejects absolute path.

not rated 7repo changed 8d ago A scan 71 tokens original Apache-2.0

Good fit Setting up llm-d, checking or customizing its deployment, and creating resources such as namespaces, storage, routes, or Helm releases when requested.

Compare 6 skills from other repositories ↓
View source ↗ llm-d-incubation/llm-d-skills
Install

Getting it into your agent

It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.

Clone the repo
git clone --depth 1 https://github.com/llm-d-incubation/llm-d-skills
agentmods
npx agentmods add skills/llm-d-incubation/llm-d-skills/deploy-llm-d

Made for: Claude Code, Codex.

Its marketplace also offers this one on its own, as the plugin deploy-llm-d — /plugin install deploy-llm-d after adding the marketplace above.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for deploy-llm-d

README.md
[![agentmods](https://agentmods.dev/badge/skills/llm-d-incubation/llm-d-skills/deploy-llm-d/github.svg)](https://agentmods.dev/skills/llm-d-incubation/llm-d-skills/deploy-llm-d)
Your own site
<a href="https://agentmods.dev/skills/llm-d-incubation/llm-d-skills/deploy-llm-d"><img src="https://agentmods.dev/badge/skills/llm-d-incubation/llm-d-skills/deploy-llm-d/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for deploy-llm-d

Your own site · 80×15
<a href="https://agentmods.dev/skills/llm-d-incubation/llm-d-skills/deploy-llm-d"><img src="https://agentmods.dev/badge/skills/llm-d-incubation/llm-d-skills/deploy-llm-d.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 71 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 6,152 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00071 $0.06152
Opus 5.5 $0.00028 $0.02461
Sonnet 5.5 $0.00014 $0.01230
Haiku 4.5 $0.00007 $0.00615

Measured 8d ago against content hash 2974d542ee92, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

deploy-llm-d scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 8d ago.

The scan reads SKILL.md. This mod also ships 1 executable file (scripts/list-guide-steps.py), listed below but not scanned — reading those needs a real analyzer, not pattern matching.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

skills/deploy-llm-d/SKILL.md · 296 lines

How it starts

The opening of the file, as written. The whole thing — 296 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Deploy llm-d Stack

Command Execution Notice

Before executing any command, the agent will:

  1. Explain what the command does and the expected outcome.
  2. Show the exact command that will be executed.
  3. Explain why the command is needed at this point of the workflow.

Rules

  1. Notify the user before creating anything. Before creating any resource (namespace, Secret, PVC, Helm release, Gateway, HTTPRoute, file, or any other Kubernetes object), state what will be created and why.
  2. Do not change cluster-level definitions. All changes are made inside the target namespace. Never modify ClusterRoles, ClusterRoleBindings, StorageClasses, Nodes, or any resource outside the target namespace. Scope every kubectl apply, helm install, and kubectl delete command with -n ${NAMESPACE}. The only exception is a guide's own cluster-scoped prerequisite step (prerequisites.gaie, prerequisites.crds, prerequisites.lws, prerequisites.keda, prerequisites.kueue, prerequisites.compute_domain). These install CRDs or controllers that the guide cannot run without. Check whether they are already present (Step 3). If they are missing, show the exact upstream command, state that it is cluster-scoped, and run it only after the user explicitly confirms. Otherwise hand the command to the cluster administrator and stop.
  3. Do not modify any existing code you did not create. Never edit files committed in the llm-d checkout (guide.yaml, README.md, values files, kustomize overlays). Customizations live in new files under deployments/ (Step 4.4).
  4. Do not send inference requests without approval. Readiness checks use Kubernetes state only. Requests populate the KV cache and change benchmark baselines; the connectivity test is opt-in (Step 5).

Overview

Since llm-d v1.0 every foundation and model guide under ${LLMD_PATH}/guides/ is a pair of files:

  • guide.yaml: the machine-readable source of truth. It declares env.static variables (NAMESPACE, ACCELERATOR_TYPE, MODEL_SERVER, INFRA_PROVIDER, MODEL, ...), a support matrix, and step lists under prerequisites, deploy, verify, optionally benchmark, and cleanup. Steps carry when: filters (selected by variable values) and skip_in: tags.
  • README.md: human-readable prose whose bash blocks are rendered from guide.yaml.

Read the full file on GitHub · 296 lines

Files

What ships with it

4 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 8d ago Changed · -132 lines · +25 tokens per session 2974d542ee92
  2. 2mo ago First seen · 428 lines · 46 tokens per session scan A 152139b74609

Subscribe to this mod's changes

deploy-llm-d is a skill published in the GitHub repository llm-d-incubation/llm-d-skills (7 stars, last pushed 8d ago), licensed Apache-2.0. It adds 71 tokens to every session and 6,152 once invoked, about $0.0003 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

CloudContainersDevOps and cloud
Related

deploy-llm-d compared with skills from other repositories

deploy-llm-d next to 6 skills from other repositories
ModTypeStarsLast pushLicenceScanTokensWorks with
deploy-llm-dllm-d-incubation/llm-d-skillsSkill78d agoApache-2.0A71Claude Code, Codex
openshell-cliNVIDIA/OpenShellSkill21kyesterdayApache-2.0A134Claude Code, Codex
azd-deploymentsickn33/agentic-awesome-skillsSkill49kyesterdayMITA29Claude Code
cdf-orchestration-setupNVIDIA/skillsSkill3.9kyesterdayApache-2.0A105Codex
langbot-deploylangbot-app/LangBotSkill19kyesterdayApache-2.0B104Claude Code, Codex
compute-env-setupcompanion-inc/feynmanSkill10k6d agoMITA45Claude Code, Codex
agent-bomLeoYeAI/openclaw-master-skillsSkill2.2k6d agoMITA99Claude Code, Codex