geniml

Machine learning toolkit for genomic interval (BED) data; use it when you need to tokenize BED collections and train embeddings for regions/cells/labels, build consensus peak universes, or run similarity search and downstream ML on chromatin accessibility datasets.

Quality

80%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Low

Low-risk findings worth noting

Quality

Content

68%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 well-organized, largely executable skill body that demonstrates the two main end-to-end pipelines and pushes detail to reference files. Main weaknesses are missing code examples for several advertised features and absent validation checkpoints in the batch workflows.

Suggestions

Add a brief validation checkpoint after tokenization (e.g. confirm the tokens/ folder is non-empty and report the token count) so the batch workflow has an explicit verify step.

Provide a short executable snippet for BEDspace training, including how the StarSpace binary path is supplied, since it is a headline feature with no code example.

Convert the flat reference list into per-feature inline links (e.g. '**BEDspace**: See references/bedspace.md') to improve navigation and progressive disclosure.

Dimension	Reasoning	Score
Conciseness	The body is organized into tight bulleted sections and code blocks with mostly domain-specific explanations, but a few prose passages (e.g. the word2vec 'document/word' analogy) could be trimmed.	4 / 5
Actionability	Region2Vec and scEmbed pipelines give copy-paste-ready code with imports and parameters, but BEDspace, Text2BedNN, BEDshift, and BBClient are described only in prose without executable examples.	4 / 5
Workflow Clarity	The build-universe → tokenize → train → evaluate sequence is clearly numbered, but the batch tokenization step lacks any validation checkpoint (e.g. verify output token counts), so the batch-operation cap applies.	3 / 5
Progressive Disclosure	The overview is well-structured and defers detail to real one-level-deep reference files, but the references appear as a flat blockquote list rather than per-feature inline links, slightly reducing navigability.	4 / 5
	Total	15 / 20 Passed

Description

92%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 pairs a concrete capability list with explicit 'use when' triggers and a well-scoped niche. The only gap is slightly thinner coverage of natural synonyms and file extensions.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete actions — 'tokenize BED collections', 'train embeddings for regions/cells/labels', 'build consensus peak universes', 'run similarity search' — giving comprehensive coverage rather than generic phrasing.	5 / 5
Completeness	Explicitly states what it does ('Machine learning toolkit for genomic interval (BED) data...') and when to use it ('use it when you need to tokenize BED collections...'), answering both with concrete trigger phrases.	5 / 5
Trigger Term Quality	Strong domain keywords ('BED', 'embeddings', 'consensus peak universes', 'similarity search', 'chromatin accessibility') that genomics users would say, but it omits file extensions (e.g. '.bed') and common synonyms like ChIP-seq/ATAC-seq.	4 / 5
Distinctiveness Conflict Risk	The niche is sharply defined — genomic interval BED ML with named sub-tools (Region2Vec, BEDspace, scEmbed) — so it is unlikely to trigger for unrelated skills.	5 / 5
	Total	19 / 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: aipoch/medical-research-skills
Path: scientific-skills/Data Analysis/geniml/SKILL.md
Commit: 63c61d3

Reviewed: 9 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.