CtrlK
BlogDocsLog inGet started
Tessl Logo

hugging-science

Use when the user is doing AI/ML work in a scientific domain such as biology, chemistry, physics, astronomy, climate, genomics, materials, medicine, ecology, energy, engineering, math, drug discovery, protein design, weather modeling, theorem proving, single-cell, or PDE solving. Hugging Science is a curated catalog of scientific datasets, models, blog posts, and interactive Spaces. This skill helps discover and use resources via `datasets`, `transformers`, the HF Inference API, `gradio_client`, and methodology citations.

71

Quality

86%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

SecuritybySnyk

Low

Low-risk findings worth noting

SKILL.md
Quality
Evals
Security

Quality

Content

78%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.

Well-structured content: a navigable five-step workflow with executable commands and clean progressive disclosure to real reference files. The main lever for improvement is conciseness and inlining a bit more executable per-resource guidance, since several sections explain concepts Claude largely already knows.

Suggestions

Tighten the 'A few important things to remember' and intro paragraphs to remove restatements of catalog value and basic explanations (e.g., why streaming helps) that Claude already knows, improving conciseness toward anchor 5.

Inline one small executable snippet per resource type (e.g., a one-line datasets load_dataset + streaming example and a one-line gradio_client call) in the 'Use the resource' step so the body is more copy-paste ready rather than delegating all code to references.

Add an explicit verify checkpoint after loading a dataset/model (e.g., 'print split sizes / sample one record before bulk processing') to push workflow_clarity toward the anchor-5 feedback-loop pattern.

DimensionReasoningScore

Conciseness

The body is mostly lean — a five-step workflow, targeted code snippets, and pointed reminders — with only minor over-explanation (e.g., restating the value of curation and explaining streaming's purpose) that could be trimmed, placing it at 'efficient; minor instances of over-explanation' rather than the fully lean anchor 5.

4 / 5

Actionability

Provides executable commands (fetch_catalog.py invocations), a runnable python-dotenv snippet, exact catalog URLs, and named reference files per resource type; it is mostly executable, though the core per-resource API usage lives in bundled references rather than inline, leaving minor gaps versus the copy-paste-ready anchor 5.

4 / 5

Workflow Clarity

A clear five-step sequence (identify domain, fetch, pick, use, cite) with decision criteria and an explicit checkpoint to present top candidates to the user before proceeding; no destructive/batch cap applies, but explicit verify-after-load feedback loops are only implicit, so it is 'clear sequence with most checkpoints, minor gaps' rather than 5.

4 / 5

Progressive Disclosure

The SKILL.md is an overview that signals one-level-deep references to real bundled files (references/using-datasets.md, using-models.md, using-spaces.md, topics-and-slugs.md, flagship-resources.md, scripts/fetch_catalog.py), with a 'Bundled resources' index — matching the clear-overview anchor 5.

5 / 5

Total

17

/

20

Passed

Description

95%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, well-formed description: it leads with a comprehensive 'Use when' trigger list, states a clear 'what' anchored to a curated scientific catalog, and names the concrete tools used. Minor specificity gap only in that 'discover and use resources' is somewhat abstract compared to fully enumerated actions.

DimensionReasoningScore

Specificity

Names the catalog surface (datasets, models, blog posts, Spaces) and concrete tools for discovery/use (datasets, transformers, HF Inference API, gradio_client, methodology citations); actions like 'discover and use' remain slightly abstract rather than fully enumerated, so it sits at 'several specific actions, minor gaps' rather than 5.

4 / 5

Completeness

Explicitly answers both 'what' (curated catalog of scientific datasets/models/blog posts/Spaces; helps discover and use resources) and 'when' (the 'Use when the user is doing AI/ML work in a scientific domain such as...' clause with concrete trigger phrases), mirroring the anchor-5 example.

5 / 5

Trigger Term Quality

The 'Use when...' clause enumerates a comprehensive, natural set of domain terms users actually say (biology, chemistry, physics, astronomy, climate, genomics, materials, medicine, ecology, energy, engineering, math, drug discovery, protein design, weather modeling, theorem proving, single-cell, PDE solving), matching the comprehensive-coverage anchor.

5 / 5

Distinctiveness Conflict Risk

A clear niche (scientific-domain ML on Hugging Face) with distinct triggers (named scientific domains and named models/tools) yields minimal conflict risk with other skills.

5 / 5

Total

19

/

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.

Validation16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository
K-Dense-AI/scientific-agent-skills
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.