CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

compute-env-setup

Prepare reproducible setup instructions and validate a user-managed named software environment on an Open-Science SSH Compute Host, including direct SSH and Slurm hosts. Use when a remote job needs packages, modules, cache variables, or a repeatable activation that the host does not already provide.

Quality

86%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Passed

No findings from the security scan

SKILL.md
Quality
Evals
Security

Quality

Content

88%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

The body is a strong, action-dense instruction set: complete executable code for probing, activation, validation, and recording, with an explicit validation feedback loop and careful ownership and removal safety. Its only real weaknesses are modest verbosity in the intro and recording sections and minor organization opportunities in a single long file.

DimensionReasoningScore

Conciseness

The body is dense, imperative, and full of Open-Science-specific contracts Claude does not already know (activation path resolution, ownership boundaries, job-failure semantics), so most tokens earn their place. Minor over-explanation remains: the three-paragraph intro (Slurm-accounting tangent, brand-path migration rules) and the long epistemic-hygiene paragraph in the recording section could be tightened.

4 / 5

Actionability

Every stage has complete, executable JavaScript (listHosts selection, batched read-only probe with an exact shell command, direct-SSH witness, Slurm submitJob with #SBATCH directives, knowledge append) plus a copy-paste conda activation file with || return $? guards. The few placeholders (<provider-known-partition>, <name>, <date>) are explicitly justified by the instruction to read host knowledge first, and both direct SSH and Slurm common cases are covered.

5 / 5

Workflow Clarity

The four sections give a clear sequence (define environment, prepare user-owned install/removal, validate where jobs run, record reusable facts), validation is a dedicated section with concrete witness commands for both host types, and an explicit failure feedback loop ('When validation fails, diagnose the layer identified by the error... Prepare revised user-run commands'). Removal safety requires the user to verify the exact canonical path first, so the destructive-operation cap does not apply.

5 / 5

Progressive Disclosure

No bundle files exist, and the four well-headed sections with no buried references give good structure and easy navigation. Minor gaps keep it below the top anchor: a long unheaded intro precedes the first section, and the per-system environment guidance (Conda/Modules/Apptainer) plus the evidence-scope policy are inlined in a ~190-line file where a reference split could help.

4 / 5

Total

18

/

20

Passed

Description

78%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

The description is strong on completeness, pairing a concrete capability statement with an explicit and specific 'Use when' trigger clause. Its main weaknesses are limited action coverage (two verbs) and some missing natural synonyms for environment tooling, which leave specificity and trigger coverage a notch below the top anchors.

Suggestions

List the additional concrete actions the skill performs (e.g., prepare user-run installation and removal plans, record validated host facts) to raise specificity from two actions to several.

Add natural tool synonyms users would say, such as conda, venv, container image, or 'install packages on the compute host', to broaden trigger term coverage.

DimensionReasoningScore

Specificity

The description names the domain ('user-managed named software environment on an Open-Science SSH Compute Host, including direct SSH and Slurm hosts') and exactly two concrete actions ('Prepare reproducible setup instructions' and 'validate' the environment). It matches the 1-2-concrete-actions anchor rather than the 4 anchor, which requires several specific actions.

3 / 5

Completeness

It explicitly answers both what ('Prepare reproducible setup instructions and validate a user-managed named software environment...') and when ('Use when a remote job needs packages, modules, cache variables, or a repeatable activation that the host does not already provide') with concrete trigger phrases, matching the top anchor's what+when pattern.

5 / 5

Trigger Term Quality

'packages, modules, cache variables', 'remote job', 'Slurm', and 'SSH' are natural phrasings users would say when they need this skill. Common variations such as conda, venv, container, cluster, or 'install' are missing, keeping it just below comprehensive coverage.

4 / 5

Distinctiveness Conflict Risk

'user-managed named software environment' and 'repeatable activation' carve a clear niche with minimal conflict risk, but the package/module/remote-job trigger territory overlaps a closely related sibling remote-compute or Slurm skill, matching 'mostly distinct; minor overlap risk' rather than the clear-niche 5 anchor.

4 / 5

Total

16

/

20

Passed

Validation

100%

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository

aipoch/open-science

Path

resources/skills/compute-env-setup/SKILL.md

Reviewed

Table of Contents

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.