← 返回 Skills 市场
pauldelavallaz

Multishot UGC

作者 Paul de Lavallaz · GitHub ↗ · v1.0.1
cross-platform ⚠ suspicious
1075
总下载
5
收藏
2
当前安装
2
版本数
在 OpenClaw 中安装
/install multishot-ugc
功能描述
Generate 10 perspective/angle variations from a single image for multi-shot UGC videos. ✅ USE WHEN: - Have a hero image and need camera angle variations - Creating multi-scene UGC videos (need different shots) - Want close-ups, wide shots, side angles from one source - Building a video with scene changes ❌ DON'T USE WHEN: - Don't have a hero image yet → use morpheus-fashion-design first - Need completely different scenes/locations → use Morpheus multiple times - Just need one image → skip this step - Want to edit images manually → use nano-banana-pro INPUT: Single image (person with product) OUTPUT: 10 PNG variations with different perspectives TYPICAL PIPELINE: Morpheus → multishot-ugc → select best 4 → veed-ugc each → Remotion edit
安全使用建议
Before installing or running this skill: - Expect to provide a COMFY_DEPLOY_API_KEY environment variable (the script will fail without it). The registry metadata failing to list this is an inconsistency you should ask the publisher to fix. - Understand that local images will be uploaded to https://api.comfydeploy.com and outputs downloaded back; do not upload sensitive or private images unless you trust ComfyDeploy and have reviewed their terms/privacy. - Verify the deployment_id and the ComfyDeploy domain independently (publisher/source is unknown and no homepage is provided). If you don't recognize the provider, request provenance or use an alternative with known trust. - Limit the API key scope if possible, store it securely, and rotate/revoke it if exposed. - If you need stronger assurance, ask the skill author to (a) declare COMFY_DEPLOY_API_KEY in the registry metadata, (b) provide a homepage or source provenance, and (c) explain data retention/processing on ComfyDeploy for uploaded images.
功能分析
Type: OpenClaw Skill Name: multishot-ugc Version: 1.0.1 The `scripts/generate.py` file contains a critical vulnerability that allows arbitrary local file upload and potential data exfiltration. The `--image` argument, intended for an image path, is read and uploaded to `api.comfydeploy.com` without validation of its content or type. An attacker could specify a path to any sensitive file (e.g., `/etc/passwd`, `~/.ssh/id_rsa`), leading to its exfiltration. Additionally, the `download_images` function is vulnerable to path traversal if the `filename` returned by the external API is malicious (e.g., `../../malicious.sh`), potentially allowing arbitrary file writes. These are significant vulnerabilities, but without clear evidence of intentional malicious design by the skill developer, it is classified as suspicious rather than malicious.
能力评估
ℹ Purpose & Capability
The SKILL.md and scripts implement a multishot image-generation flow that queues a ComfyDeploy deployment and downloads outputs — this aligns with the described purpose. However the registry metadata lists no required environment variables or primary credential while the script requires a COMFY_DEPLOY_API_KEY; that metadata omission is a mismatch.
ℹ Instruction Scope
Instructions and the included script only perform actions relevant to the task: optionally upload a local image, queue a deployment, poll for completion, and download images. SKILL.md does not disclose that an API key is required (the script does), which is an important omission for runtime behavior.
✓ Install Mechanism
No install spec is provided (instruction-only with small helper script and pyproject). There is a declared Python dependency (requests) in pyproject.toml, but nothing is automatically downloaded or executed by an install step in the registry metadata.
⚠ Credentials
The code requires a single environment secret (COMFY_DEPLOY_API_KEY) to call api.comfydeploy.com; that credential is proportionate to the stated purpose. The concern is that the skill metadata did not declare this required env var or a primary credential, which is a visibility/consent issue and may surprise users. No unrelated credentials are requested.
✓ Persistence & Privilege
The skill does not request permanent 'always' presence, does not modify other skills or system-wide settings, and is user-invocable only. It runs network calls to the declared third-party API but does not attempt privilege escalation or persistent system changes.
如何使用
  1. 确保已安装 OpenClaw(本地或 Docker 部署)
  2. 在对话框中输入安装命令:/install multishot-ugc
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /multishot-ugc 触发
  4. 根据 Skill 的参数说明提供必要输入,即可获得结构化输出
版本历史
v1.0.1
Improved routing: added negative examples and pipeline info
v1.0.0
Initial release: Generate 10 perspective variations of an image for UGC video production using ComfyDeploy's MULTISHOT-UGC workflow.
元数据
Slug multishot-ugc
版本 1.0.1
许可证 —
累计安装 2
当前安装数 2
历史版本数 2
常见问题

Multishot UGC 是什么?

Generate 10 perspective/angle variations from a single image for multi-shot UGC videos. ✅ USE WHEN: - Have a hero image and need camera angle variations - Creating multi-scene UGC videos (need different shots) - Want close-ups, wide shots, side angles from one source - Building a video with scene changes ❌ DON'T USE WHEN: - Don't have a hero image yet → use morpheus-fashion-design first - Need completely different scenes/locations → use Morpheus multiple times - Just need one image → skip this step - Want to edit images manually → use nano-banana-pro INPUT: Single image (person with product) OUTPUT: 10 PNG variations with different perspectives TYPICAL PIPELINE: Morpheus → multishot-ugc → select best 4 → veed-ugc each → Remotion edit. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 1075 次。

如何安装 Multishot UGC?

在 OpenClaw 或 Claude Code 对话框中运行命令「/install multishot-ugc」即可一键安装,无需额外配置。

Multishot UGC 是免费的吗?

是的,Multishot UGC 完全免费(开源免费),可自由下载、安装和使用。

Multishot UGC 支持哪些平台?

Multishot UGC 跨平台运行,可在任意部署了 OpenClaw / Claude Code 的环境中使用(cross-platform)。

谁开发了 Multishot UGC?

由 Paul de Lavallaz(@pauldelavallaz)开发并维护,当前版本 v1.0.1。

💬 留言讨论