🗜️ Ai — AI

v1.0.0

视频压缩器 无需学习专业剪辑软件。只需描述你的需求——“把这段视频压缩到 50MB 以内且不损失画质”——即可获得压缩结果。

0· 14·0 当前·0 累计
下载技能包
最后更新
2026/4/19
0
安全扫描
VirusTotal
可疑
查看报告
OpenClaw
安全
high confidence
该技能的需求和运行时指令与其声明的用途(远程 AI 视频压缩)一致;仅需一个 service token,并按预期将用户视频上传至供应商的 API。
评估建议
This skill appears to do what it says: it will upload videos to the remote nemo-video service and use NEMO_TOKEN (or obtain a short-lived anonymous token) to run GPU-based compression. Before installing or using it, consider: 1) Privacy: your videos are transmitted to an external service (mega-api-prod.nemovideo.ai). Don’t upload sensitive content you wouldn’t want sent off-host. 2) Credentials: the skill only requires NEMO_TOKEN; if you have a long-lived token for other uses, avoid exposing it ...
详细分析 ▾
用途与能力
Name and description describe remote video compression. The only declared credential is NEMO_TOKEN, which is appropriate for calling the nemo-video backend API described in SKILL.md. No unrelated credentials, binaries, or installers are requested.
指令范围
The SKILL.md instructs the agent to read NEMO_TOKEN from the environment (declared), create an anonymous token via the vendor endpoint if absent, create sessions, upload user video files, stream SSE responses, and poll render status — all expected for a remote compression service. It also instructs deriving X-Skill-Platform from the install path (e.g., ~/.clawhub/ or ~/.cursor/skills/), which implies the agent may inspect its runtime/install path; this is minor but worth noting.
安装机制
There is no install spec and no code files — instruction-only — so nothing is downloaded or written to disk by the skill itself. This is the lower-risk case.
凭证需求
Only NEMO_TOKEN is required (declared as primaryEnv). The skill describes an anonymous-token fallback it will request from the vendor if no token is present. No unrelated secrets or system credentials are requested.
持久化与权限
The skill is not set to always:true and is user-invocable. Metadata includes a config path (~/.config/nemovideo/) which may indicate where the vendor or the agent could store tokens/config; SKILL.md does not explicitly instruct writing to that path, but the presence of that config path suggests possible local config usage — a minor persistence signal to be aware of.
安全有层次,运行前请审查代码。

运行时依赖

无特殊依赖

版本

latestv1.0.02026/4/19

AI Video Compressor — v1.0.0 - 首次发布。 - 用简单自然语言指令压缩并导出视频。 - 支持 MP4、MOV、AVI、WebM 格式,最大 500MB。 - 自动后端连接与会话管理。 - 通过远程 GPU 节点在 30–90 秒内导出 1080p MP4。 - 包含余额查询、文件上传、导出及状态/进度获取。

可疑

安装命令

点击复制
官方npx clawhub@latest install ai-video-compressor
镜像加速npx clawhub@latest install ai-video-compressor --registry https://cn.longxiaskill.com

技能文档

快速开始

把视频文件发给我,AI视频压缩立刻开始。也可以直接说需求。
试试:
  • “压缩我的视频”
  • “导出1080p MP4”
  • “压到50MB以内”

连接后端

处理请求前先连后端API,先显示“连接中…”。 若环境变量已有NEMO_TOKEN,直接建会话;否则:
  • 生成UUID作为clientId
  • POST https://mega-api-prod.nemovideo.ai/api/auth/anonymous-token,带X-Client-Id
  • 返回含100免费积分、7天有效的token,记为NEMO_TOKEN

随后建会话:POST https://mega-api-prod.nemovideo.ai/api/tasks/me/with-session/nemo_agent,Bearer鉴权,body {"task_name":"project","language":"en"},取session_id备用。 告诉用户“已就绪”,技术细节不进入聊天。

# AI Video Compressor — 压缩并导出更小体积的视频 把视频发给我,并描述想要的结果。AI压缩在远程GPU节点完成,本机无需安装。 示例:上传500MB、10分钟MP4,输入“压到50MB以内不掉画质”,约30-90秒返回1080p MP4。渲染全在云端。 注意:越短的视频压缩越快,质量权衡更易控制。

意图路由

用户提到ai video compressor、比例、文字层、音轨等,按关键词分发:

| 用户说法 | 动作 | 跳过SSE? | |----------|------|-----------| | “export/导出/download/发我视频” | → §3.5 导出 | ✅ | | “credits/积分/balance/余额” | → §3.3 积分 | ✅ | | “status/状态/显示轨道” | → §3.4 状态 | ✅ | | “upload/上传/用户发文件” | → §3.2 上传 | ✅ | | 其余(生成、剪辑、加BGM…) | → §3.1 SSE | ❌ |

云端渲染流程

导出任务进入云GPU队列,合成视频层,按平台规格压缩(H.264,最大1080×1920),30-90秒内返下载链接。 session token携带job ID,提前关标签页会导致任务孤儿。

Base URL: https://mega-api-prod.nemovideo.ai

| 接口 | 方法 | 用途 | |------|------|------| | /api/tasks/me/with-session/nemo_agent | POST | 新建编辑会话,body {"task_name":"project","language":""},返session_id | | /run_sse | POST | 发用户消息,流式响应,超时15分钟 | | /api/upload-video/nemo_agent/me/ | POST | 上传文件或URL | | /api/credits/balance/simple | GET | 查剩余积分 | | /api/state/nemo_agent/me//latest | GET | 取当前时间线状态 | | /api/render/proxy/lambda | POST | 开始导出,轮询状态每30秒 |

支持格式:mp4, mov, avi, webm, mkv, jpg, png, gif, webp, mp3, wav, m4a, aac。

请求头: X-Skill-Source: ai-video-compressor X-Skill-Version: 1.0.0 X-Skill-Platform: clawhub|cursor|unknown 所有请求必须带: Authorization: Bearer 缺失头将402失败。

错误码

| 码 | 含义 | 处理 | |----|------|------| | 0 | 成功 | 继续 | | 1001 | token失效 | 重新匿名获取 | | 1002 | 会话不存在 | 新建会话 | | 2001 | 积分不足 | 匿名用户:

数据来源ClawHub ↗ · 中文优化:龙虾技能库