← Explore all skills

SPACES / Hugging Face

Gradio LoRA Space Builder

Pins Hugging Face's LoRA Space skill, which tries a public read before asking for a token and waits for one approval before a private publish.

“Reads a public LoRA without a token, asks before a private Space is created, and does not smoke-test a Space that is not running.”

01 / THE REASONING

Why this made the selection.

  • An unauthenticated read comes first. A token is requested only when no valid cached token can read a private or gated repo.
  • The three Space files are approved in one message, the Space is private by default, and the username is confirmed before creation. A smoke test waits until the runtime is RUNNING.

02 / THE REVIEW RECORD

What we actually inspected.

Source review has boundaries.
A clear record is more useful than a “safe” badge.

Material inspected

  • skills/huggingface-lora-space-builder/SKILL.md
  • README.md
  • LICENSE

Our findings

  • The file says to try reading the LoRA repo without a token, and not to ask for a token before checking a cached login.
  • If the base model is outside the Qwen-Image, LTX, and Krea 2 notes, it says to ask whether to proceed by analogy or stop, rather than guessing.
  • app.py, requirements.txt, and README.md are written and then approved together before publishing.
  • The Space defaults to private, and the target username is confirmed before create_repo.
  • torch.compile is forbidden because it is incompatible with ZeroGPU.
  • gradio info is not run unless the Space is RUNNING, and a green build is not a reason to skip the inference check.
  • A 401 or 403 on upload means the token lacks write scope.

Not established by this review

  • The resource was not executed. No Hub repo was read, no Space was created, and no token was supplied.
  • gradio info and gradio predict were not run.
  • The references directory was not opened.

The review applies to the material and revision named here. A newer upstream release can change its behavior.

03 / PUT IT TO WORK

Add a skill to your workflow.

Upstream setup instructions ↗
  1. The repository README says client marketplaces expose hf-cli. The printed plugin line is `/plugin install hf-cli@huggingface/skills` after `/plugin marketplace add huggingface/skills`. That does not install this skill.
  2. The same README shows `hf skills add <skill-name>` for other skills. The angle-bracket token is a placeholder, so it is not recorded as the install command.
  3. This review did not install the skill.

Before you start

  • A Hugging Face account when a Space is later published
  • A write-scoped token only in the case the skill describes: no cached token can read a private or gated LoRA, or upload returns 401 or 403

04 / FOLLOW THE EVIDENCE

The source trail.

Our notes are separate from the original resource.
Check upstream before adopting a new version.

  1. https://github.com/huggingface/skills/blob/c3ff942be3e07116b475253f0a8fc4f54682c1a8/skills/huggingface-lora-space-builder/SKILL.md

    Supports: unauthenticated read first, one approval before publish, private default, and no smoke test until RUNNING

  2. Repository README ↗Checked

    https://github.com/huggingface/skills/blob/c3ff942be3e07116b475253f0a8fc4f54682c1a8/README.md

    Supports: hf-cli plugin install and the placeholder hf skills add line

  3. Repository LICENSE ↗Checked

    https://github.com/huggingface/skills/blob/c3ff942be3e07116b475253f0a8fc4f54682c1a8/LICENSE

    Supports: Apache License 2.0

Evidence & Ask