CtrlK
BlogDocsLog inGet started
Tessl Logo

modal-serverless-gpu

Run approved CPU or GPU work through OpenScience compute_job 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.

72

Quality

88%

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

93%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.

A tight, highly actionable skill body with concrete examples and a clear workflow. The only gap is the absence of an explicit validation/retry checklist, which keeps workflow_clarity just below full marks.

Suggestions

Add an explicit validation checkpoint in the workflow, e.g. after dispatch poll status before reporting, and a 'If status is failed, read logs, fix inputs, and re-dispatch with a new job ID' retry step to make the feedback loop explicit.

DimensionReasoningScore

Conciseness

The body is lean and dense with rules, assumes Claude's competence, and avoids explaining basics like what Modal or a sandbox is; every line carries operational guidance.

5 / 5

Actionability

A complete, copy-paste-ready JSON example shape with concrete command, target, uploads, artifacts, packages, gpu, and resources covers the common dispatch case directly.

5 / 5

Workflow Clarity

A clear five-step sequence with an inspection checkpoint ('plan' vs 'start') and an error-recovery loop for transient control-plane interruptions is present, but there is no fully explicit validate -> fix -> retry checklist.

4 / 5

Progressive Disclosure

At under 50 lines with no need for external references, the well-organized Contract / Workflow / Example shape sections satisfy the simple-skill exception for full marks.

5 / 5

Total

19

/

20

Passed

Description

83%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.

A specific, well-scoped description that names concrete capabilities and a clear niche. Its main weakness is the trigger phrasing, which uses 'Use for...' use-case listing rather than natural 'Use when.../when the user mentions...' trigger terms.

Suggestions

Replace 'Use for...' with an explicit trigger clause such as 'Use when the user asks to run CPU or GPU scientific scripts on Modal, or needs isolated compute provisioning, durable outputs, logs, status, cancellation, or recovery.'

Add natural synonyms a user would actually say (e.g. 'run a job on Modal', 'GPU sandbox', 'serverless compute') so the skill surfaces for conversational phrasings, not just internal 'compute_job' terminology.

DimensionReasoningScore

Specificity

The description lists multiple concrete capabilities — 'isolated scientific scripts, dependency provisioning, durable outputs, logs, status, cancellation, and recovery' — alongside the core action, giving comprehensive coverage.

5 / 5

Completeness

It clearly states the 'what' and gives a 'Use for...' trigger list of use cases, but lacks an explicit 'Use when...' / 'when the user mentions...' phrasing, so the 'when' could be more concrete.

4 / 5

Trigger Term Quality

Natural keywords like 'CPU or GPU work', 'scientific scripts', and 'Modal' are present, but 'OpenScience compute_job' is internal jargon and common synonyms/variations are thin.

4 / 5

Distinctiveness Conflict Risk

The niche is sharp — Modal via OpenScience compute_job for scientific GPU/CPU work — and the 'Never invoke the Modal SDK or CLI directly' boundary reinforces a distinct trigger with minimal conflict risk.

5 / 5

Total

18

/

20

Passed

Validation

93%

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

Validation15 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

frontmatter_unknown_keys

Unknown frontmatter key(s) found; consider removing or moving to metadata

Warning

Total

15

/

16

Passed

Repository
synthetic-sciences/openscience
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.