← 返回 Skills 市场
aqbjqtd

Ffmpeg Master Pro

作者 sniper-one · GitHub ↗ · v2.3.1 · MIT-0
cross-platform ⚠ suspicious
306
总下载
0
收藏
2
当前安装
8
版本数
在 OpenClaw 中安装
/install ffmpeg-master-pro
功能描述
FFmpeg Master Pro - 全能视频处理技能。当用户需要视频处理时使用,支持视频转换、视频压缩、视频编辑等。当用户要求视频处理, 视频转换, 视频压缩, FFmpeg, 视频编辑, 视频转码, 视频剪辑, 字幕处理, 视频优化, 批量视频处理, GIF转换, 视频翻转, 速度调节, 音频提取, 视频合...
使用说明 (SKILL.md)

环境要求

  • 必装ffmpegffprobe(建议从 ffmpeg.org 或 via winget/choco 安装)
  • 可选nvidia-smi(NVIDIA 加速)、vainfo(Intel QSV)、amdgpu-info(AMD 加速)

首次使用前验证:ffmpeg -versionffprobe -version

执行建议

执行方式

FFmpeg 处理推荐使用 exec background:true 避免长时间阻塞会话。示例:

exec background:true command:"ffmpeg -i input.mp4 -c:v libx264 -crf 23 output.mp4"

完成后通知(可选)

任务完成后推荐使用 message 工具通知用户:

✅ 视频处理完成
文件:output.mp4 | 操作:转码压缩

注意:仅发送文件名即可,避免暴露完整路径。

文件路径处理

  • 输入/输出路径直接使用用户提供的位置,不做跨分区复制
  • Windows 路径(如 C:\Users\...)和 Unix 路径(如 /mnt/c/...)均支持
  • 临时文件处理:如有需要可使用系统临时目录,任务结束后清理

核心工作流

所有工作流的完整命令、参数说明和代码示例见 详细工作流

转码与优化

# 工作流 触发词 详细文档
1 智能转码与压缩 压缩、转码、格式转换、减小文件
2 智能参数优化 自动识别6种视频类型并优化编码
3 预设模板系统 YouTube/B站/微信/抖音等平台预设
4 精确文件大小控制 指定目标大小,两遍编码偏差\x3C5%

剪辑与分析

# 工作流 触发词 详细文档
5 Smart Cut 混合剪辑 剪辑、裁剪、cut
6 关键帧分析 关键帧、最佳剪辑点

编辑与特效

# 工作流 触发词 详细文档
7 字幕处理 提取/嵌入/烧录字幕
8 滤镜与特效 水印、旋转、翻转
9 GIF 转换 GIF、动图
10 翻转与镜像 镜像、左右/上下翻转

变换与合并

# 工作流 触发词 详细文档
11 速度调节 加速、慢动作、倍速
12 音频提取与移除 提取音频、去音频、静音
13 视频合并 合并、拼接、concat
14 宽高比调整 16:9、9:16、竖屏、横屏

智能决策系统

技能会自动执行以下决策,无需手动干预:

  • 内容类型识别:自动检测电影/动漫/屏幕录制/体育/音乐视频/老旧视频,选择最优编码参数。详见 优化指南
  • GPU 加速:自动检测 NVIDIA NVENC / AMD AMF / Intel QSV 并优先使用
  • 质量验证:自动校验输出文件的分辨率、时长、码率,可选 VMAF/SSIM/PSNR 评估

预设模板

内置 8+ 平台预设(assets/presets/):youtube、bilibili、wechat、douyin、social_media、archival、preview、web_optimized。使用时提及平台名称即可自动匹配。

参考文档

工作流程

  1. 分析用户需求,匹配触发词到对应工作流
  2. ffprobe 分析输入视频(格式、编码、分辨率、码率、时长)
  3. 如有脚本可用(scripts/),优先执行脚本;否则按详细工作流文档构建 ffmpeg 命令
  4. 执行转换,验证输出质量
  5. 批量任务参考 detailed_workflows.md 附录中的批量处理器
安全使用建议
What to consider before installing: - Functional fit: The skill's functionality (building/running ffmpeg commands, GPU detection, presets, batch processing) matches its description — this looks like a legitimate FFmpeg helper. - Declared requirements mismatch: The metadata does not list required binaries or config paths, but SKILL.md and the code clearly expect ffmpeg and ffprobe (and optionally nvidia-smi/amdgpu-info/vainfo). If your environment lacks these tools the skill will fail; the missing declaration is an oversight you should ask the author to correct. - Local execution risks: The skill's scripts execute subprocesses (ffmpeg/ffprobe and system GPU tools) and read/write files. If the agent constructs command lines from untrusted user input, there is a risk of command injection or unintended access to local files. Prefer running this skill in a controlled/sandboxed environment until you audit how inputs are sanitized. - Files written to your system: The PresetManager will create ~/.config/ffmpeg-master/presets for user presets. Expect the skill to create temporary output files and logs in temporary directories or in user-specified locations. If you want to avoid changes to your home directory, run in a VM or container. - No remote downloads detected: There is no installer that downloads arbitrary code from the network at install time, which reduces supply-chain risk. Still review the bundled scripts for any network calls before enabling autonomous invocation. - Recommended actions before enabling: 1) Ask the author/maintainer to correct metadata to declare required binaries and config paths. 2) Review the scripts that will run (scripts/), paying attention to subprocess invocations and any code that touches the network or unsanitized inputs. Look for any subprocess.run or os.system uses that concatenate untrusted strings. 3) Test the skill in an isolated environment (container/VM) with non-sensitive files to verify behavior. 4) If you will run it on a multi-user or production machine, enforce least privilege (run under a restricted user, restrict read/write directories) and confirm input sanitization. If you want, I can highlight specific files/functions that build command lines or perform subprocess calls so you can inspect them for unsafe string handling.
功能分析
Type: OpenClaw Skill Name: ffmpeg-master-pro Version: 2.3.1 The ffmpeg-master-pro skill bundle is a comprehensive and well-structured toolkit for advanced video processing. It implements sophisticated logic for video analysis, intelligent bitrate calculation, and hardware acceleration detection. The Python scripts (e.g., decision_engine.py, ffmpeg_builder.py, and two_pass_encoder.py) use best practices for executing external commands, specifically employing subprocess.run and subprocess.Popen with shell=False and shlex.split to mitigate shell injection risks. The documentation and SKILL.md instructions are strictly aligned with the tool's purpose and even include security-conscious advice, such as avoiding the exposure of full file paths to the user. No indicators of data exfiltration, malicious persistence, or unauthorized network activity were found.
能力评估
Purpose & Capability
The code and SKILL.md implement a comprehensive FFmpeg-based processing toolkit (transcode, smart-cut, presets, GPU detection). That matches the skill name/description. However the registry metadata claims no required binaries while SKILL.md and the code require ffmpeg/ffprobe (and optionally nvidia-smi/amdgpu-info/vainfo). The missing declaration of those required binaries in metadata is an inconsistency.
Instruction Scope
Runtime instructions explicitly tell the agent to run ffmpeg/ffprobe and to 'exec background:true' for long tasks. The SKILL.md directs the agent to prefer executing included scripts under scripts/ when available. The included Python scripts call subprocess.run on ffmpeg/ffprobe/nvidia-smi/other tools and read/write files. The SKILL.md and code accept and use user-provided file paths directly (input/output), but give no rigorous guidance about sanitizing paths or validating untrusted inputs. This combination increases risk of command-line injection or accidental exposure of local files if user inputs are not safely handled.
Install Mechanism
There is no install spec that downloads remote archives or runs third-party installers; the skill ships many code files with the bundle. No network download URLs or extract steps were found in the provided metadata. Because nothing is pulled from an external URL during install, the install mechanism itself is low risk — but the presence of many executable scripts means runtime execution (by the agent) is the primary operational surface.
Credentials
requires.env and primary credential are empty in the registry, but SKILL.md and code require local binaries (ffmpeg/ffprobe) and optionally commands that reveal hardware (nvidia-smi, amdgpu-info, vainfo). The PresetManager writes to a user config directory (~/.config/ffmpeg-master/presets) and creates files in the user's home directory — this config path usage is not declared in required config paths. No secrets are requested, but the skill will access local system state and create files in the user's profile without having declared those config paths.
Persistence & Privilege
always:false (no forced inclusion). The skill writes/creates a user presets directory under the home config path and may create temporary files during processing. It does not request persistent elevated privileges or modify other skills. Persisting presets under ~/.config is reasonable for a local tool but should have been declared as a config path; users should be aware of this file creation.
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 ffmpeg-master-pro
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /ffmpeg-master-pro 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
v2.3.1
修复可疑模式问题
v2.3.0
Add mandatory background execution rules + cross-platform notification via message tool (no channel parameter)
v2.2.2
修复 api_reference.md 中失效链接
v2.2.1
修复 api_reference.md 中失效链接
v2.2.0
按 skill-creator 标准重构 SKILL.md:消除内容重复,精简为导航索引
v2.1.0
重组文档编号,修复链接锚点,补充缺失内容
v1.1.0
修复:添加运行时依赖声明(ffmpeg/ffprobe/python3),修复frontmatter格式,移除非标准字段
v1.0.0
首次发布:14个工作流覆盖视频处理全场景,含智能转码压缩、Smart Cut剪辑、GIF转换、速度调节、音频提取、视频合并、宽高比调整等,内置8+平台预设模板,支持GPU加速
元数据
Slug ffmpeg-master-pro
版本 2.3.1
许可证 MIT-0
累计安装 2
当前安装数 2
历史版本数 8
常见问题

Ffmpeg Master Pro 是什么?

FFmpeg Master Pro - 全能视频处理技能。当用户需要视频处理时使用,支持视频转换、视频压缩、视频编辑等。当用户要求视频处理, 视频转换, 视频压缩, FFmpeg, 视频编辑, 视频转码, 视频剪辑, 字幕处理, 视频优化, 批量视频处理, GIF转换, 视频翻转, 速度调节, 音频提取, 视频合... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 306 次。

如何安装 Ffmpeg Master Pro?

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

Ffmpeg Master Pro 是免费的吗?

是的,Ffmpeg Master Pro 完全免费,采用 MIT-0 许可证,可自由下载、安装和使用。

Ffmpeg Master Pro 支持哪些平台?

Ffmpeg Master Pro 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 Ffmpeg Master Pro?

由 sniper-one(@aqbjqtd)开发并维护,当前版本 v2.3.1。

💬 留言讨论