hugging-face-datasets

Create and manage datasets on Hugging Face Hub. Supports initializing repos, defining configs/system prompts, streaming row updates, and SQL-based dataset querying/transformation. Designed to work alongside HF MCP server for comprehensive dataset workflows.

Quality

66%

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

Fix and improve this skill with Tessl

tessl review fix ./backend/cli/skills/data-engineering/hugging-face-datasets/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.

The content is highly actionable with abundant executable examples, but it is padded with basic SQL primers and repeated command demos, and inline reference material that could be split into separate files keeps progressive disclosure from its top score.

Suggestions

Remove the 'Useful DuckDB SQL Functions' primer covering basics Claude already knows (COUNT/SUM/AVG, GROUP BY, LENGTH, LOWER/UPPER) and keep only the non-obvious bits (hf:// path syntax, USING SAMPLE RESERVOIR).

Deduplicate the init/config/add_rows command blocks that appear in both 'Recommended Workflow' and 'Commands Reference'.

Add an explicit validation checkpoint to destructive/batch workflows (e.g., 'describe' the result or check stats before/after --push-to and add_rows) to lift workflow clarity.

Dimension	Reasoning	Score
Conciseness	Most content is concrete and earns its place, but the 'Useful DuckDB SQL Functions' section explains basics Claude already knows (COUNT/SUM/AVG, GROUP BY, LOWER/UPPER, array indexing) and several command examples (init, config, add_rows) are repeated across sections.	2 / 3
Actionability	The body is packed with copy-paste-ready, fully executable commands and a complete Python API block (e.g., 'uv run scripts/sql_manager.py query --dataset "cais/mmlu" --sql "..."'), matching the executable anchor.	3 / 3
Workflow Clarity	Sequences exist (Recommended Workflow, numbered combined examples), but destructive/batch operations like --push-to and add_rows lack explicit validate-then-retry checkpoints, capping workflow clarity at 2 per the rubric.	2 / 3
Progressive Disclosure	Sections are well-organized and the two referenced scripts (scripts/sql_manager.py, scripts/dataset_manager.py) are real files, but the full SQL syntax reference, template structures, and DuckDB function primer all live inline in SKILL.md rather than in separate one-level-deep reference docs.	2 / 3
	Total	9 / 12 Passed

Description

67%

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

The description is specific and carves out a clear, distinct niche, but it omits explicit "Use when..." trigger guidance, capping completeness and trigger-term quality below their maxima.

Suggestions

Add an explicit 'Use when...' clause listing natural trigger phrases (e.g., 'Use when creating, editing, or querying Hugging Face datasets, or when the user mentions HF Hub, dataset repos, or SQL over datasets').

Include common keyword variations users might say (e.g., 'HF Hub', 'Hugging Face datasets', 'dataset rows', 'query a dataset with SQL') to broaden trigger coverage.

Dimension	Reasoning	Score
Specificity	The description lists multiple concrete actions — "initializing repos", "defining configs/system prompts", "streaming row updates", and "SQL-based dataset querying/transformation" — matching the multi-action anchor.	3 / 3
Completeness	The "what" is clearly answered, but there is no "Use when..." clause or equivalent explicit trigger guidance, which per the rubric caps completeness at 2.	2 / 3
Trigger Term Quality	It includes natural domain terms ("Hugging Face Hub", "datasets", "SQL-based dataset querying") but lacks common variations and any "Use when..." trigger phrasing, so it stops short of full coverage.	2 / 3
Distinctiveness Conflict Risk	The Hugging Face dataset niche is specific, and the explicit "Designed to work alongside HF MCP server" framing carves out a distinct boundary, making conflicts unlikely.	3 / 3
	Total	10 / 12 Passed

Validation

81%

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 — 13 / 16 Passed

Validation for skill structure

Criteria	Description	Result
skill_md_line_count	SKILL.md is long (561 lines); consider splitting into references/ and linking	Warning
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning
referenced_paths_exist	Referenced path issues: 2 missing	Warning

	Total	13 / 16 Passed

Repository: synthetic-sciences/openscience
Path: backend/cli/skills/data-engineering/hugging-face-datasets/SKILL.md
Commit: 52845c3

Reviewed: about 11 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.