CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

hugging-face-cli

Execute Hugging Face Hub operations using the `hf` CLI. Use when the user needs to download models/datasets/spaces, upload files to Hub repositories, create repos, manage local cache, or run compute jobs on HF infrastructure. Covers authentication, file transfers, repository creation, cache operations, and cloud compute.

71

Quality

89%

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

76%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.

The body is highly actionable — dense with executable `hf` commands and realistic worked patterns — and lean in style. Its weaknesses are structural: the inline command listings duplicate both the quick-reference table and the bundled reference file, and workflows involving destructive operations (repo/file deletion, cache removal) lack any verification steps.

Suggestions

Add validation checkpoints around destructive and batch operations: e.g., run `hf auth whoami` before uploads, confirm the target with `hf models info <repo_id>` before `hf repo delete`/`hf repo-files delete`, and verify results after `hf upload` or `hf cache rm`.

De-duplicate the body: either drop the inline "Core Commands" listings (keeping the Quick Command Reference table plus Common Patterns) or shrink them to a few key examples, leaving the full command surface to references/commands.md as the References section already advertises.

Turn the Publish Model and Sync Space patterns into explicit numbered sequences with a verification step between each (e.g., confirm the repo exists before uploading, confirm the tag landed with `hf repo tag list`).

DimensionReasoningScore

Conciseness

Lean, command-first content with essentially no explanation of concepts Claude already knows, but the "Quick Command Reference" table duplicates nearly every command that is restated verbatim in the "Core Commands" section — a redundant block that could be trimmed. Not 5 because of that duplication; not 3 because padding is confined to one section.

4 / 5

Actionability

Fully executable, copy-paste-ready commands throughout, including concrete instances with real repo IDs (meta-llama/Llama-3.2-1B-Instruct, MiniMaxAI/MiniMax-M2.1) that cover the common cases across download, upload, repo management, cache, and jobs.

5 / 5

Workflow Clarity

Multi-step patterns exist (Publish Model: create → upload → tag; Sync Space; endpoint deploy), but destructive and batch operations (hf repo delete, hf repo-files delete, hf cache rm, hf endpoints delete --yes) have no validation or verification steps; the only checkpoint is the HF_TOKEN check. Per the guideline, a destructive/batch skill without validation is capped at 3.

3 / 5

Progressive Disclosure

Well-organized sections with clearly signaled, one-level-deep references (references/commands.md and references/examples.md both exist and contain no further nesting). Not 5 because the ~110-line inline "Core Commands" section substantially duplicates the 954-line references/commands.md, going beyond "a few key examples inline, bulk in separate file".

4 / 5

Total

16

/

20

Passed

Description

100%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 description: concrete, comprehensive, third-person, with an explicit "Use when…" trigger clause covering all major capability areas of the CLI. The closing "Covers authentication, file transfers, repository creation, cache operations, and cloud compute" sentence is mildly redundant with the trigger clause but does not hurt clarity.

DimensionReasoningScore

Specificity

Lists multiple concrete actions — "download models/datasets/spaces, upload files to Hub repositories, create repos, manage local cache, or run compute jobs" — giving comprehensive coverage of the CLI's capabilities in third-person voice, matching the top anchor rather than having minor gaps.

5 / 5

Completeness

Explicitly answers both: what ("Execute Hugging Face Hub operations using the `hf` CLI" with enumerated actions) and when ("Use when the user needs to download models/datasets/spaces, upload files to Hub repositories, create repos, manage local cache, or run compute jobs"), with concrete trigger phrases.

5 / 5

Trigger Term Quality

Natural phrases users would say ("download models", "upload files", "create repos", "manage local cache", "run compute jobs") plus "Hugging Face Hub" and "hf CLI" cover the common variations comprehensively; no common term is missing.

5 / 5

Distinctiveness Conflict Risk

Clear niche (Hub operations via the `hf` CLI) with distinct triggers tied to Hugging Face-specific vocabulary; minimal conflict risk with other skills.

5 / 5

Total

20

/

20

Passed

Validation

87%

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

Validation — 14 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

metadata_version

'metadata.version' is missing

Warning

frontmatter_unknown_keys

Unknown frontmatter key(s) found; consider removing or moving to metadata

Warning

Total

14

/

16

Passed

Repository

synthetic-sciences/openscience

Path

backend/cli/skills/llm-tools/hugging-face-cli/SKILL.md

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.