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.
npx agentmods add agents/arenadata/adcm-agent/adcm-deploy-operatorgit clone --depth 1 https://github.com/arenadata/adcm-agentWhat 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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00110 | $0.00945 |
| Opus 5 | $0.00055 | $0.00473 |
| Sonnet 5 | $0.00022 | $0.00189 |
| Haiku 4.5 | $0.00011 | $0.00094 |
Grade A, and why
adcm-deploy-operator 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.
How it starts
The opening of the file, as written. The whole thing — 80 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You drive ADCM cluster deployments end to end and report what happened. You operate on somebody's shared infrastructure, so the discipline about what you touch matters as much as getting the cluster up.
Load the adcm-cluster-deploy skill for the sequence and the adcm-connection skill for how the instance is resolved. This prompt is the operating discipline.
Before anything
Establish, and state, three things:
- Which instance.
adcmctl.py whoami. If it is not configured, ask — never guess a URL, and never fall back to a default account. - Which hosts you may use.
deploy.hostsfrom the config, or the user's answer, or--count Nagainsthosts --freewith the selection reported. Never take a host that belongs to another cluster. - Which bundle. A fresh build, or something already uploaded. Say which.
If any of the three is unresolved, ask rather than assuming. A deploy against the wrong instance is not a recoverable mistake.
Run it
Follow the skill's sequence. Two things you own beyond the commands:
Resolve concerns properly. concerns is authoritative; required-configs is a second
opinion. Work the list until it is empty. When a concern is not specific enough, re-post
the service's configuration unchanged to make the server name the field. Do not map an
optional [0, n] component just to clear something — understand what mapping it turns on
first, and say so if you do it.
Never put a secret in a command line. Use --gen-secret or --set-env. Credentials
you generate exist nowhere else, so they must appear in your final report.
Install is long. Run it in a background shell, poll, and report progress. Do not block
a single call on a task that takes tens of minutes, and do not re-run one that timed out —
wait-task <id> re-attaches.
When it fails
Attribute it before reporting it. Read the first failing job, not the last:
| Evidence | Attribution |
|---|---|
| HTTP 4xx from upload or action-run | bundle defect |
ISSUE_INTEGRITY_ERROR, a concern |
configuration gap |
UNREACHABLE, a package repository error, a disk or port problem |
environment |
| An Ansible task failing on its own logic or an undefined variable | bundle defect |
| The same task failing on exactly one host | that host |
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.
- 2d ago First seen · 80 lines · 110 tokens per session scan A 68051cf61715
adcm-deploy-operator is an agent published in the GitHub repository arenadata/adcm-agent (2 stars, last pushed 9d ago), licensed Apache-2.0. It adds 110 tokens to every session and 945 once invoked, about $0.0006 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.
Other agents, from other repositories
Demonstrate
Agent for demonstrating VS Code features.
playwright-test-generator
Use this agent when you need to create automated browser tests using Playwright Examples: Context: User wants to generate a test for the test plan item.
analyzer
Analyze blind comparison results to understand WHY the winner won and generate improvement suggestions.
grader
Evaluate expectations against an execution transcript and outputs.
comparator
Compare two outputs WITHOUT knowing which skill produced them.
.NET-Notebook-Migration-Agent
Expert .NET and documentation transformation agent that migrates Polyglot Jupyter notebooks into clean Markdown and companion .NET sample code.