Skip to content
Back to skills

Model Card

ASecurity

Write an HF-compatible model card (data, classes, metrics, limits) next to a checkpoint.

  • 8 stars
  • 0 votes
  • 0 copies
  • 1 view
  • Added September 19, 2026
ai-agentsgit

Security analysis

A100/100

Scanned September 19, 2026

npx -y skills add Adilmunawar/ZD-claude-plugin --skill model-card --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Model Card?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Model Card
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/adilmunawar-model-card/badge)](https://www.skillsdirectory.com/skills/adilmunawar-model-card)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: model-card
disable-model-invocation: true
argument-hint: "<checkpoint-dir>"
description: Write an HF-compatible model card (data, classes, metrics, limits) next to a checkpoint.
---

Write `model_card.md` (HF-compatible YAML header + body):

- **Task / classes** (list with ids), **input** (sensor, bands, resolution, normalisation), **AOI & seasons** in training data, **label source** and count.
- **Architecture** (lib, encoder, params), **training** (epochs, LR schedule, loss, augmentation, hardware, time), **git SHA**.
- **Metrics**: per-class IoU/F1 table on val and test; boundary F1 if applicable; confusion matrix image path.
- **Known failure modes** (e.g. small fields < 0.2 ha, cloud shadow, mixed cropping), **intended use**, **not for**.
- **How to run inference** (3-line snippet) and **checkpoint format** (dict keys).
Then print the path. Push to HuggingFace only on explicit request.

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…