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 skills add TianchenGuan/duke-cs-cluster-skill --skill duke-clustergit clone --depth 1 https://github.com/TianchenGuan/duke-cs-cluster-skillWrote 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.
[](https://agentmods.dev/skills/tianchenguan/duke-cs-cluster-skill/duke-cluster)<a href="https://agentmods.dev/skills/tianchenguan/duke-cs-cluster-skill/duke-cluster"><img src="https://agentmods.dev/badge/skills/tianchenguan/duke-cs-cluster-skill/duke-cluster/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.
<a href="https://agentmods.dev/skills/tianchenguan/duke-cs-cluster-skill/duke-cluster"><img src="https://agentmods.dev/badge/skills/tianchenguan/duke-cs-cluster-skill/duke-cluster.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00064 | $0.03923 |
| Opus 5 | $0.00032 | $0.01962 |
| Sonnet 5 | $0.00013 | $0.00785 |
| Haiku 4.5 | $0.00006 | $0.00392 |
Grade A, and why
duke-cluster scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 11d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 250 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Duke CS Compute Cluster
Skill version: 3
Applies to: Duke Computer Science Slurm cluster (login.cs.duke.edu)
Last verified: 2026-04-13 via sinfo, scontrol show partition, sacctmgr, module avail, 12 smoke jobs.
Regenerate with: bash scripts/verify_skill.sh and update tables.
Docs: https://computing.cs.duke.edu — subpages: Compute/{cluster, slurm, gpu-computing}, Storage/disk_usage, Software/{environment-modules, apptainer}.
Pre-Flight Checklist
Every job script should do all of the following. Copy from templates/ — don't reconstruct from memory.
set -euo pipefail— fail fast on errors, undefined vars, broken pipes.- Set cache dirs explicitly (never rely on inheritance):
export HF_HOME=/usr/xtmp/$USER/.cache/huggingface export TORCH_HOME=/usr/xtmp/$USER/.cache/torch export PIP_CACHE_DIR=/usr/xtmp/$USER/.cache/pip - Always specify
--mem— the default is 4.5 GB, too low for any real work. - Always specify
--time— the default is 4 days, which blocks scheduling. - Point
--output/--errorto/usr/xtmp/%u/...or$HOME— never/tmp(node-local, invisible from login node). - Log GPU info:
nvidia-smi -L 2>/dev/null || echo 'none'
Partitions
Only compsci and compsci-gpu are generally accessible. Other partitions
(grisman, wiseman, nlplab, nlplab-core, bhuwan, rudin, skynet, wills) are
lab-specific — submit with -A PARTITIONNAME -p PARTITIONNAME.
| Partition | Purpose | Default time | Max time | Nodes | Total CPUs | GPUs? | Default mem |
|---|---|---|---|---|---|---|---|
compsci |
CPU work | 4 days | 90 days | 48 | 2,448 | No | 4.5 GB |
compsci-gpu* |
GPU work | 4 days | 90 days | 62 | 3,192 | Yes (217 total) | 4.5 GB |
* compsci-gpu is the default partition. No debug partition exists — use --time=00:05:00 for quick smoke tests. No preemption (PreemptMode=OFF).
What ships with it
15 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.
- references/apptainer.md 970 B
- references/distributed_training.md 2.1 KB
- references/failure_modes.md 5.3 KB
- references/node_ips.md 7.8 KB
- scripts/check_cluster_load.sh 1.2 KB runs code
- scripts/check_gpu_contention.sh 2.4 KB runs code
- scripts/check_my_jobs.sh 2.1 KB runs code
- scripts/estimate_wait.sh 1.1 KB runs code
- scripts/prestage_hf_model.sh 1.4 KB runs code
- scripts/verify_skill.sh 1.9 KB runs code
- templates/cpu_only.sbatch 656 B
- templates/multi_gpu_single_node.sbatch 988 B
- templates/multi_node.sbatch 1.2 KB
- templates/single_gpu.sbatch 912 B
- templates/smoke_test_gpu.sbatch 556 B
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.
- 11d ago First seen · 250 lines · 64 tokens per session scan A 18b74d32f04d
duke-cluster is a skill published in the GitHub repository TianchenGuan/duke-cs-cluster-skill (6 stars, last pushed 5mo ago), licensed MIT. It adds 64 tokens to every session and 3,923 once invoked, about $0.0003 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 skills, from other repositories
latchbio-integration
Build, register, debug, and operate bioinformatics workflows on Latch using the Python SDK, CLI, Latch Data and Registry, Nextflow, Snakemake, programmatic execution, and Latch MCP. Use when authoring or deploying Latch workflows, configuring resources or interfaces, moving data, integrating Registry, or launching and…
remote-compute-ssh
Evaluate and use SSH Remote Compute before choosing where to run GPU, high-memory, parallel, batch, model-inference, bioinformatics, or other long-running scientific work; supports short remote commands and asynchronous jobs with automatic harvest and analysis.
modal-compute
Run explicitly chosen research benchmark or replication jobs on Modal's serverless infrastructure. Use when a Feynman research workflow needs burst remote GPU compute and the Modal CLI is available.
dnanexus-integration
DNAnexus cloud genomics platform. Build apps/applets, manage data (upload/download), dxpy Python SDK, run workflows, FASTQ/BAM/VCF, for genomics pipeline development and execution.
latchbio-integration
Latch platform for bioinformatics workflows. Build pipelines with Latch SDK, @workflow/@task decorators, deploy serverless workflows, LatchFile/LatchDir, Nextflow/Snakemake integration.
modal-serverless-gpu
Run approved CPU or GPU work through OpenScience computejob on the user's configured Modal account. Use for isolated scientific scripts, dependency provisioning, durable outputs, logs, status, cancellation, and recovery. Never invoke the Modal SDK or CLI directly.