可信 — 风险评分 5/100
上次扫描:1 天前 重新扫描
5 /100
ffmpeg_multimedia_editing
使用 FFmpeg 进行多媒体编辑,包括视频剪辑、拼接、转码、特效、截图、水印、变速、音频处理等 20 种操作
Legitimate FFmpeg wrapper skill with no malicious indicators - subprocess calls to ffmpeg/ffprobe are the declared core functionality, all processing is local, no network access or credential harvesting.
技能名称ffmpeg_multimedia_editing
分析耗时34.3s
引擎pi
可以安装
This skill is safe to use. It performs exactly as documented - local multimedia processing using FFmpeg binaries.
资源类型声明权限推断权限状态证据
文件系统 WRITE WRITE ✓ 一致 SKILL.md declares file write capability; code writes processed media to outputs/…
命令执行 WRITE WRITE ✓ 一致 SKILL.md declares ffmpeg/ffprobe as required binaries; common.py line 168: subpr…
网络访问 NONE NONE No network requests found in codebase
环境变量 READ READ ✓ 一致 Only reads OUTPUT_ROOT environment variable for output directory

目录结构

23 文件 · 75.8 KB · 2446 行
Python 21f · 2337L Markdown 1f · 105L YAML 1f · 4L
├─ 📁 agents
│ └─ 📋 openai.yaml YAML 4L · 593 B
├─ 📁 scripts
│ ├─ 🐍 audio_concat.py Python 104L · 3.3 KB
│ ├─ 🐍 audio_denoise.py Python 83L · 2.4 KB
│ ├─ 🐍 audio_extract_replace.py Python 94L · 2.8 KB
│ ├─ 🐍 audio_volume.py Python 85L · 2.5 KB
│ ├─ 🐍 common.py Python 441L · 13.8 KB
│ ├─ 🐍 video_compress_scale.py Python 97L · 3.1 KB
│ ├─ 🐍 video_concat.py Python 83L · 2.3 KB
│ ├─ 🐍 video_convert.py Python 101L · 3.2 KB
│ ├─ 🐍 video_crop_rotate.py Python 98L · 2.6 KB
│ ├─ 🐍 video_firstlast_frame.py Python 77L · 2.0 KB
│ ├─ 🐍 video_gif_convert.py Python 105L · 3.2 KB
│ ├─ 🐍 video_image_compose.py Python 105L · 3.3 KB
│ ├─ 🐍 video_pip.py Python 76L · 2.2 KB
│ ├─ 🐍 video_reverse.py Python 72L · 1.8 KB
│ ├─ 🐍 video_screenshot.py Python 131L · 4.2 KB
│ ├─ 🐍 video_speed.py Python 88L · 2.4 KB
│ ├─ 🐍 video_split.py Python 77L · 2.1 KB
│ ├─ 🐍 video_subtitle.py Python 103L · 3.2 KB
│ ├─ 🐍 video_transition.py Python 95L · 2.9 KB
│ ├─ 🐍 video_trim.py Python 93L · 2.7 KB
│ └─ 🐍 video_watermark.py Python 129L · 4.3 KB
└─ 📝 SKILL.md Markdown 105L · 4.8 KB

安全亮点

✓ All subprocess calls are to declared binaries (ffmpeg, ffprobe) - no arbitrary command execution
✓ No network requests or external API calls - all processing is local as documented
✓ No credential harvesting or sensitive path access
✓ No obfuscation techniques (base64, eval, etc.)
✓ Clean codebase with no IOCs or suspicious patterns
✓ Standard logging to local files only
✓ Input validation prevents path traversal attacks