SkillFoxx

MiniMax H3 Cases and Guides skill

Three Codex and Claude Code skills: search MiniMax H3 cases and full prompts, pick tutorial routes and prepare submissions

  • 220 ★ on the repository
  • Last commit
  • MIT
Skill

Minimal access

Why this level

  • The reading skills call the hosted public catalog over the network
All reasons and checks

skynotsilent/awesome-minimax-h3-casesAuthor:SkyNotSilent

Install

In your terminal, with SkillFoxx CLI

npx skillfoxx add skills/minimax-h3-cases-and-guides

Detects the agents on your machine, shows the access level and pins the version.

Other ways to install

Access and checks

How we review

Minimal accessThe tool runs nothing and makes no network calls.

  • The reading skills call the hosted public catalog over the network

This is third-party code. Review the repository files before installing.

What it does

MiniMax H3 Cases and Guides is a public library of videos made in MiniMax H3 (also known as Hailuo H3), with the source and creator of each case. The repository adds three Codex and Claude Code skills that read the hosted catalog. minimax-h3-prompt-library finds real cases and returns the full prompt when the source published it; it never generates or rewrites prompts. minimax-h3-tutorial-guide picks a checked route by OS, GPU, VRAM, time and goal without guessing commands or versions. minimax-h3-submission-helper prepares a submission of a case or tutorial and does not claim to verify or publish it. The two reading skills include standalone query clients for the catalog and fail explicitly when it is unavailable. Each guide on the site has a Copy for AI button that builds an execution package for an agent and requires checking the current README before any commands.

Who it is for. Video creators and prompt authors working with MiniMax H3 who want to find ready cases and checked guides from Codex or Claude Code.

Good fit when

  • A scene needs a real MiniMax H3 case with a fully published prompt
  • You need a launch route for your OS, GPU and VRAM
  • You want to prepare a submission of your case or tutorial to the library
  • You need cases by genre, duration or style among hundreds of options

Not a fit when

  • You need a prompt the source did not publish: the skill never invents or translates one
  • You work offline: the reading skills call the hosted catalog and report an error when it is unavailable
  • You want video generation itself: the repository gathers cases and guides and does not run the model

Example request

Find a ten-second cityscape case in the MiniMax H3 library and return the full prompt if the source published it

Limitations

The reading skills work only while the hosted catalog is reachable. A prompt is returned only when the source published it in full, otherwise the skill says so. The submission skill does not confirm verification or publication; maintainers decide. The repository is large (about 600 MB with assets), so install only the skills you need via npx skills add with their names without cloning it all.

How to disable. Delete the minimax-h3-prompt-library, minimax-h3-tutorial-guide and minimax-h3-submission-helper folders from the Codex and Claude Code skills directories.

FAQ

What is MiniMax H3?

A video generation model also searched as Hailuo H3, Hailuo 3.0 and 海螺 H3. The library collects playable examples, full prompts and practical guides for it.

Does the skill invent prompts?

No. minimax-h3-prompt-library returns a prompt only when the source published it in full and never generates, rewrites or translates one itself.

How do I choose a launch route?

minimax-h3-tutorial-guide selects a checked route by OS, GPU, VRAM, time and goal. It does not guess missing commands, versions or compatibility.

What does Copy for AI do?

Builds an execution package for an agent from a guide's structured fields. The package requires checking the upstream's current README and forbids guessed commands and compatibility claims.

How do I submit my case?

Through minimax-h3-submission-helper, which prepares an Issue-ready submission. The project maintainers accept, verify and publish; the skill claims no such authority.

Discussion

No questions or reviews yet.

Skills

The skill retells a reference video, waits for your confirmation and writes MiniMax H3 prompts per scene and shot.

SkillExtended access325
Official

The official mmx CLI of the MiniMax platform with an agent skill: text, images, video, speech, audio transcription, vision and search

CLIExtended access2.2K

Codex skill that makes short MiniMax H3 videos from your images on a Google Colab GPU, with a job queue and result download

SkillExtended access136

The skill turns a short idea into a full MiniMax-H3-Image prompt and can render the picture through the MiniMax API.

SkillExtended access312

A library of 280-plus skills and annotated cases for writing short-video prompts in MiniMax H3 and Seedance 2.0

SkillMinimal access331

A director skill carries a short drama from script to a ready MiniMax H3 prompt package with H1 to H36 checks.

SkillExtended access71

Comparison with similar tools

Cards that share the main search query, and how they differ.

ToolOfficialRepository starsMain difference
MiniMax H3 Cases and Guides (this card)no220
MiniMax H3 video reverseno325breaks a finished video into a description, while this searches a library of cases and tutorials
MiniMax-H3-Image Prompt Writingno312writes image prompts, while this one returns published video prompts

README in short

The README follows scenarios: watch cases, find complete prompts, learn by goal and hardware, share your work. Then come the three skills with hard boundaries, data and automation, local run and contribution rules. The case and prompt counters at the top update automatically.

README badge

Original SKILL.md

---
name: minimax-h3-prompt-library
description: This skill should be used when the user asks to "find a MiniMax H3 case", "show the original prompt", "look up a Hailuo 3.0 example", "verify whether a prompt was published", or "retrieve source-attributed H3 examples". It retrieves real cases and verbatim public prompts only; it refuses prompt generation, rewriting, reconstruction, reverse engineering, translation, or workflow decomposition.
---

# MiniMax H3 Case and Public Prompt Lookup

Retrieve source-attributed MiniMax H3 / Hailuo 3.0 video cases from the catalog. Return prompt text only when its provenance and completeness fields support that the full public text is retained.

Keep this skill lookup-only. Never create a new prompt, adapt an existing prompt, combine prompt fragments, reconstruct a missing prompt, reverse-engineer a video, or decompose a hidden production workflow.

## Source of truth

Use the packaged `scripts/query.mjs` client to read the public catalog. It uses `H3_LIBRARY_URL` when configured and otherwise reads the official hosted catalog. Treat each original `sourceUrl` as the attribution authority. If the network or catalog is unavailable, report that lookup failed; never substitute invented cases.

```bash
node <installed-skill-directory>/scripts/query.mjs "city cinematic"
```

When working inside the repository, `data/cases.json` remains the publishing source of truth. Valid category, style, and scene keys are defined in `data/taxonomy.json`.

Use case metadata to search by model, mode, category, style, scene, creator, title, or visible result. Present localized labels to users. Keep editorial summaries separate from source-published prompt text.

## Retrieval workflow

1. Identify the requested model, mode, scene, style, creator, or result.
2. Run the packaged query client for the closest factual matches. Resolve its path relative to this installed skill directory, not the user's current project. Do not assume repository data files are installed.
3. Return the case title, short factual summary, model, mode, author, catalog identifier, and original source URL.
4. Check `promptProvenance` before returning prompt text.
5. Return a complete prompt only when `promptCompleteness` is `complete` (or omitted for legacy records) and provenance is `official-verbatim` or `creator-verbatim`.
6. Never return a Prompt whose text is truncated or ends as an incomplete excerpt; treat it as unavailable.
7. Treat `unknown`, `official-adapted`, `reconstructed`, or any unrecognized provenance as unavailable. Do not present stored text under those labels as an original prompt.
8. Preserve public prompt wording, order, punctuation, timestamps, formatting, and language exactly. Do not clean, translate, shorten, expand, or normalize it.

## Output contract

Return these sections:

```text
CASE MATCHES
PUBLIC PROMPT
SOURCE AND PROVENANCE
DISCLOSURE
```

Report a problem