adcm-deploy

A command for deploying a test cluster through Arenadata Cluster Manager from a prepared bundle.

In plain words
What is it for?
Use it to upload a bundle and create a test cluster, including selecting hosts and handling the required cluster prototype and licence steps.
Why use it?
It guides the deployment process and requires the target system, available hosts, and bundle to be identified before making changes.

Command

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add commands/arenadata/adcm-agent/adcm-deploy
Clone the repo
git clone --depth 1 https://github.com/arenadata/adcm-agent
Per session 13 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 716 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
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 $0.00013 $0.00716
Opus 5 $0.00006 $0.00358
Sonnet 5 $0.00003 $0.00143
Haiku 4.5 $0.00001 $0.00072

Measured 2d ago against content hash effde91f12b0, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

adcm-deploy 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 2d ago.

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.

commands/adcm-deploy.md · 67 lines

How it starts

The opening of the file, as written. The whole thing — 67 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Deploy a cluster on the configured ADCM instance. Target: $ARGUMENTS

Load the adcm-cluster-deploy skill — it owns the sequence, the concern workflow and the attribution table. For a long or unattended run, delegate to the adcm-deploy-operator agent instead of driving it inline.

Establish three things before touching the instance

State each one in your first message:

  1. Which instanceadcmctl.py whoami. Unconfigured means /adcm-connect first.
  2. Which hosts you may usedeploy.hosts from the config, the user's answer, or --count N against hosts --free with the selection reported. Never take a host that belongs to another cluster; never assume a host name.
  3. Which bundle — a fresh build, or something already uploaded (bundles --mine).

If any is unresolved, ask. A deploy against the wrong instance is not recoverable.

Then run the sequence

ADCM="python3 ${CLAUDE_PLUGIN_ROOT}/scripts/adcmctl.py"

$ADCM upload                                   # if you built something
$ADCM prototype --type cluster                 # id + licence status
$ADCM accept-license <prototypeId>             # only when unaccepted
$ADCM create-cluster --prototype-id <id>
$ADCM add-services --cluster <clusterId>
$ADCM add-hosts --cluster <clusterId> --names <hosts>
$ADCM map --cluster <clusterId>
$ADCM concerns --cluster <clusterId>           # work this list until it is empty
$ADCM run-action --cluster <clusterId> --name install --wait

Two things you own beyond the commands:

Concerns are authoritative. required-configs misses fields that are invalid rather than absent. When a concern is not specific enough, re-post the service's configuration unchanged (set-config with no --set) and read the validation error.

Never put a secret in a command line. --gen-secret KEY --show-generated or --set-env KEY=ENV_VAR. Credentials you generate exist nowhere else — they must be in your report.

Install takes tens of minutes. Run it in a background shell and poll; do not block one call on the whole task, and do not re-run one that timed out — wait-task <id> re-attaches.

Read the full file on GitHub · 67 lines

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. 2d ago First seen · 67 lines · 13 tokens per session scan A effde91f12b0

Subscribe to this mod's changes

adcm-deploy is a command published in the GitHub repository arenadata/adcm-agent (2 stars, last pushed 9d ago), licensed Apache-2.0. It adds 13 tokens to every session and 716 once invoked, about $0.0001 per session on Opus 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.