ray-data

Scalable data processing for ML workloads. Streaming execution across CPU/GPU, supports Parquet/CSV/JSON/images. Integrates with Ray Train, PyTorch, TensorFlow. Scales from single machine to 100s of nodes. Use for batch inference, data preprocessing, multi-modal data loading, or distributed ETL pipelines.

Quality

75%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./backend/cli/skills/data-engineering/ray-data/SKILL.md

Quality

Content

61%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 thorough, actionable reference with good structure and genuine bundle references, weakened by duplication between the body and reference files and missing validation checkpoints for batch write/inference operations.

Suggestions

Remove the inline Integration (Ray Train/PyTorch/TensorFlow) and duplicate Transformation sections from the body, keeping only a concise example and a link to references/integration.md and references/transformations.md to eliminate redundancy.

Add a validation/verification step after batch writes and inference (e.g., re-read a sample, check row counts, or assert output schema) so batch operations include a feedback checkpoint.

Drop the marketing 'Use cases' company list and move the version-sensitive 'Ray 2.40.0+' detail into a dedicated version/deprecation note rather than the main body.

Dimension	Reasoning	Score
Conciseness	Mostly efficient code examples, but integration and transformation content is duplicated between the body and the reference files, and the marketing-style 'Use cases' (Pinterest/ByteDance/Spotify) plus the time-sensitive 'Ray 2.40.0+' add padding Claude does not need.	3 / 5
Actionability	Rich, mostly copy-paste-ready code spanning read, transform, filter, groupby, GPU, write, and framework integration, with only minor template placeholders like MyModel() and undefined clean_data/tokenize/augment.	4 / 5
Workflow Clarity	A clear pipeline pattern is shown, but batch operations (write_parquet, batch inference) lack validation or verification checkpoints, capping workflow clarity per the batch-operation rule.	3 / 5
Progressive Disclosure	Well-organized sections with real, clearly-signaled one-level-deep references (transformations.md, integration.md); minor gap is that some referenced content is also inlined in the body.	4 / 5
	Total	14 / 20 Passed

Description

88%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 strong, third-person description that concretely states capabilities and provides an explicit 'Use for...' trigger clause. Minor gains possible from adding file-extension synonyms.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete capabilities—'Streaming execution across CPU/GPU', 'supports Parquet/CSV/JSON/images', 'Integrates with Ray Train, PyTorch, TensorFlow', 'Scales from single machine to 100s of nodes'—with comprehensive coverage.	5 / 5
Completeness	Explicitly answers both what ('Scalable data processing... Streaming execution...') and when ('Use for batch inference, data preprocessing, multi-modal data loading, or distributed ETL pipelines').	5 / 5
Trigger Term Quality	Strong natural terms ('batch inference', 'data preprocessing', 'multi-modal data loading', 'distributed ETL pipelines') but lacks file-extension synonyms like '.parquet' that users might mention.	4 / 5
Distinctiveness Conflict Risk	Ray-specific distributed ML data processing niche is mostly distinct, with minor overlap risk against general ETL or pandas-centric skills.	4 / 5
	Total	18 / 20 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: synthetic-sciences/openscience
Path: backend/cli/skills/data-engineering/ray-data/SKILL.md
Commit: 3a6c3a9

Reviewed: 1 day 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.