modal-serverless-gpu

Serverless GPU cloud platform for running ML workloads. Use when you need on-demand GPU access without infrastructure management, deploying ML models as APIs, or running batch jobs with automatic scaling.

Quality

77%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

High

Do not use without reviewing

Fix and improve this skill with Tessl

tessl review fix ./09-infrastructure/modal/SKILL.md

Quality

Content

65%

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

A highly actionable, code-rich reference with good external file signaling, but it is dense and monolithic in the main body and leans on some marketing-flavored prose. Workflow validation checkpoints for stateful/deploy operations are implicit rather than explicit.

Suggestions

Move the promotional 'Key features' bullets ('Sub-second cold starts', 'pay-per-second pricing', 'scale to 100+') out of the body or condense them to one line; keep the skill instructional rather than promotional.

Split the inline reference sections (GPU configuration, container images, web endpoints, secrets, scheduling) into the existing references/ files, leaving SKILL.md as a concise overview with pointers.

Add an explicit validation checkpoint for stateful/deploy workflows — e.g. after volume writes, verify with a reload or `modal volume list` before proceeding; after deploy, confirm the endpoint responds.

Dimension	Reasoning	Score
Conciseness	Mostly lean code-and-tables reference, but it includes promotional feature bullets ('Sub-second cold starts: Rust-based infrastructure', 'pay-per-second GPU pricing without idle costs', 'scale to 100+ GPUs instantly') that describe rather than instruct and could be trimmed.	2 / 3
Actionability	Abundant copy-paste-ready executable code (hello-GPU, inference endpoint, volumes, endpoints, batching, secrets, scheduling) and concrete commands ('modal run', 'modal setup', 'modal secret create').	3 / 3
Workflow Clarity	The Quick start offers a light sequence (install → hello world → endpoint), but operations involving persistent state or deployment (volume writes, deploys) lack explicit validation/checkpoint steps; recovery guidance lives only in a static issues table.	2 / 3
Progressive Disclosure	Two real, clearly signaled one-level-deep references (advanced-usage.md, troubleshooting.md) exist, but the ~340-line body inlines GPU config, images, storage, endpoints, batching, secrets, scheduling, and debugging reference material that could be split out.	2 / 3
	Total	9 / 12 Passed

Description

90%

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 strong description that clearly answers what the skill does and when to use it, with natural trigger terms and a distinct niche. The only blemish is second-person voice ('you need') in the trigger clause.

Suggestions

Rewrite the trigger in third person to match the guideline, e.g. 'Use when the user needs on-demand GPU access...' instead of 'Use when you need on-demand GPU access...'.

Dimension	Reasoning	Score
Specificity	Lists concrete actions ('running ML workloads', 'deploying ML models as APIs', 'running batch jobs with automatic scaling'), but the second-person phrase 'Use when you need' violates the third-person guideline, reducing the score by one.	2 / 3
Completeness	Explicitly states both what it does ('Serverless GPU cloud platform for running ML workloads') and when to use it via an explicit 'Use when...' clause with three triggers.	3 / 3
Trigger Term Quality	Covers natural terms a user would say — 'GPU access', 'ML workloads', 'ML models as APIs', 'batch jobs', 'automatic scaling' — with good variation.	3 / 3
Distinctiveness Conflict Risk	Targets a clear niche (Modal serverless GPU for ML) with triggers unlikely to fire for unrelated skills.	3 / 3
	Total	11 / 12 Passed

Validation

93%

Warnings & errors only

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

Validation — 15 / 16 Passed

Validation for skill structure

Criteria	Description	Result
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	15 / 16 Passed

Repository: Orchestra-Research/AI-Research-SKILLs
Path: 09-infrastructure/modal/SKILL.md
Commit: 773a529

Reviewed: about 6 hours ago

Table of Contents

Discovery Implementation Validation

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.