minimax-h3 SKILL.md skillA
minimax-h3 from mini-yifan/open-source-video-gen-skill is an agent skill: a SKILL.md folder an AI agent loads when a task calls for it.
mdr add mini-yifan/open-source-video-gen-skill/minimax-h3@git:20260908.0a40107Pins git:20260908.0a40107 in mdr.lock. Pin the fingerprint instead.
- Audit grade
- A, 17 of 17 checks passed
- Kind
- SKILL.md skill
- Repository
- mini-yifan/open-source-video-gen-skill
- Path
skills/minimax-h3/SKILL.md- Versions recorded
- 4
- Latest version
- git:20260908.0a40107, committed Sep 8, 2026
- First version recorded
- Aug 30, 2026
- Size
- 17,033 bytes
- Repository stars
- 40
- License
- a license file GitHub does not identify
Use a remote SeetaCloud or AutoDL ComfyUI GPU instance to generate MiniMax H3 videos. Connects over SSH, discovers the current H3 workflows, helps choose U03/U06/U02, uploads references, submits a job, and downloads the result. MiniMax H3 supports 5–15 seconds per generation; choose the duration from the user's request and the scene instead of assuming a fixed default. Use when the user asks for MiniMax H3, ComfyUI, SeetaCloud, AutoDL video generation, text-to-video, image-to-video, or…
Collected from public GitHub. Every version is saved exactly as it was and never changes. Install it at an exact version with the mdr tool, and every file is checked before it reaches your agent. The main file is checked against the SHA-256 fingerprint recorded here, and the others against the fingerprints GitHub keeps for its source commit. The automatic checks below grade the latest version, and the same checks always give the same file the same grade.
What the file says
# MiniMax H3：云端视频生成 本技能负责使用远程 GPU 上的 MiniMax H3 / ComfyUI 生成视频。它提供连接、工作流发现、参数选择、素材上传、任务提交、轮询和下载能力，不替用户做不必要的导演决策。 MiniMax H3 单次生成时长为 **5–15 秒**。如果用户指定时长，严格使用用户的时长；如果用户没有指定，Agent 根据动作数量、对白长度、节奏和目标用途自行选择一个 5–15 秒的时长，并在提交前说明选择结果。不要因为技能示例、脚本默认值或过去的任务而固定使用 10 秒或 15 秒。 ## 分工 - 本技能：连接云端机器、发现当前工作流、选择生成路径、准备上传、提交任务、轮询和下载。 - [`h3-prompt-writing`](../h3-prompt-writing/SKILL.md)：编写 H3 提示词正文。本技能只整理 brief，不在这里发明提示词格式。 - [`autodl-app-instance`](../autodl-app-instance/SKILL.md)：当用户使用 AutoDL 应用实例且没有提供 SSH 时，负责开机、等待可用和关机。 不要把 SSH 密码、AutoDL Token 或其他凭据写入仓库、提示词或技能文件。 ## AutoDL 组合模式（批次开始开机，批次结束关机） 当目标是 AutoDL 应用实例，或用户明确要求“开机后生成、全部完成后关机”时，本技能必须先调用 `autodl-app-instance` 的 `boot`，再进行任何 H3 连接、上传或提交。一个用户请求中包含多个视频时，把它们视为一个 H3 批次；只有 `boot` 返回实例已 running、`COMFY_READY` 且队列可访问后，才能进入 H3 流程。 组合模式的生命周期由 AutoDL 技能管理： 1. 在批次开始前解析并记录实例 UUID，建立本批次的 job 清单。 2. 每个批次只执行一次 `boot`；捕获最新 `SEETACLOUD_BASE_URL`，在当前执行环境中显式设置。子进程里的 `export` 不会自动传回父进程。`boot` 的面板发现走"快照 service_*_domain 并行直探"，不依赖 SSH，macOS/Linux/Windows 通用（Windows 无 expect 也能跑）。 3. 在同一个生命周期内完成全部 H3 job 的 brief、提示词、上传、提交、轮询、下载与落盘秒级校验。单个视频使用一次 submit/poll 脚本，不代表要重新 boot。归一化、抽帧验收、滤镜、调参重跑等本地后处理放到关机之后，不得因此推迟关机。 …
Read the whole file at its exact version.
How to install
mdr add mini-yifan/open-source-video-gen-skill/minimax-h3@sha256:fe4237a7933df7d8The line at the top of this page pins the label git:20260908.0a40107, which is what most people want. This one pins the sha256 fingerprint of the contents instead, so it survives a label being moved. Either way the exact version is written to mdr.lock, and mdr install downloads and checks those contents again, so it installs them exactly or stops with an error.
[](https://markdownregistry.com/a/art_xafbampo2lhl5uiu)
3 badge image fetches in 30 days. What the badge shows.
Versions
The registry has recorded 4 versions of this file, between Aug 30, 2026 and Sep 8, 2026, about one every 3 days.
| version | committed | commit | size | audit | |
|---|---|---|---|---|---|
| git:20260908.0a40107 latest | 2026-09-08 | 0a40107 | 17,033 B | A | view · diff |
| git:20260904.ec83cad | 2026-09-04 | ec83cad | 16,866 B | A | view · diff |
| git:20260830.cee067c | 2026-08-30 | cee067c | 15,767 B | A | view · diff |
| git:20260830.8cf2ae7 | 2026-08-30 | 8cf2ae7 | 15,499 B | A | view |
Audit of the latest version
- pass: Frontmatter block present
- pass: Frontmatter declares a name
- pass: Frontmatter declares a description
- pass: Size between 200 bytes and 200 KB (17033 bytes)
- pass: No zero-width or bidi control characters
- pass: No instruction hidden inside an HTML comment
- pass: No link to an exfiltration or paste host
- pass: No credential-shaped string
- pass: No instruction to send local credentials anywhere
- pass: No text hidden with inline styles
- pass: No prompt-injection phrasing
- pass: No curl or wget piped into a shell
- pass: No recursive delete of root, home or parent
- pass: No instruction to read or print local credentials
- pass: No base64 blob over 200 characters
- pass: No link to a raw IP address
- pass: No script tag
Source
mini-yifan/open-source-video-gen-skill · 40 stars · license a license file GitHub does not identify · pushed 2026-09-11 · branch main
API
GET https://markdownregistry.com/api/v1/artifacts/art_xafbampo2lhl5uiu GET https://markdownregistry.com/api/v1/resolve?ref=mini-yifan/open-source-video-gen-skill/minimax-h3 GET https://markdownregistry.com/api/v1/blob/fe4237a7933df7d85979758978e8aeb3f61dc9796d25d00afb268721d0a8324b
Your AI assistant does the legwork. You hear about the deals worth your word. Hand yours the standing instructions at modelranch.com and it joins the network that reads files like this one.
Guides: the SKILL.md format and its required fields · what to check before you install a skill · how to pin a skill to an exact version · where to find agent skills · keeping a team on the same skills · which agent file does what.