← 返回 Skills 市场
mobai-6543

media-processor

作者 mobai-6543 · GitHub ↗ · v1.1.0 · MIT-0
cross-platform ⚠ suspicious
90
总下载
0
收藏
0
当前安装
2
版本数
在 OpenClaw 中安装
/install media-processor-agent
功能描述
OpenClaw 核心素材加工工具。支持对图片和视频(含 URL、m3u8、mp4)进行自动化处理,包括下载、压缩、添加水印和格式转换。
安全使用建议
This skill appears to do what it says: download and process images/videos using ffmpeg and Pillow. Before installing, consider: (1) you must have FFmpeg available on the host (or rely on static-ffmpeg from pip); (2) the skill will download arbitrary HTTP/HTTPS inputs and write temporary files to /tmp and outputs to ./data/media or the input directory — avoid using untrusted URLs or run inside a sandbox to limit risk from malicious/malformed media; (3) ffmpeg and media libraries can be vulnerable to specially crafted files, so process untrusted inputs in an isolated environment and keep ffmpeg up to date; (4) the skill does not request secrets or network callbacks, but verify your runtime environment (permissions, disk space) before use.
功能分析
Type: OpenClaw Skill Name: media-processor-agent Version: 1.1.0 The media processor skill lacks input validation on the 'input' parameter in run.py, allowing for arbitrary local file reads and potential Server-Side Request Forgery (SSRF) via remote URLs. While these capabilities are aligned with the tool's stated purpose of processing images and videos, the absence of path sanitization or URL whitelisting constitutes a security vulnerability that could be exploited to access sensitive system files (e.g., /etc/passwd) or internal network resources.
能力评估
Purpose & Capability
Name/description (media processing for images and video) align with the included code and SKILL.md: run.py implements downloading, image/video compression, watermarking, and conversion using PIL and ffmpeg-python.
Instruction Scope
SKILL.md confines runtime behavior to invoking run.py with type/action/input/output/text. The implementation will download HTTP/HTTPS inputs to /tmp or process URLs directly with ffmpeg and writes outputs to ./data/media or the input directory. This file I/O and network access are expected for the stated purpose but are worth noting (processing of arbitrary remote media is supported).
Install Mechanism
There is no install spec; it's an instruction+code skill. requirements.txt lists Pillow, ffmpeg-python, and static-ffmpeg. README correctly instructs installing system ffmpeg. No external download URLs or archive extraction are embedded in the install spec.
Credentials
The skill does not request environment variables, credentials, or config paths. It attempts to load a system font and optionally uses the static-ffmpeg Python package to adjust ffmpeg paths; both are reasonable for media processing and proportionate to the stated function.
Persistence & Privilege
always is false and the skill does not modify other skills or system-wide agent settings. It creates temporary files under /tmp and may create ./data/media for outputs — behavior consistent with its role.
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 media-processor-agent
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /media-processor-agent 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
v1.1.0
- Updated dependencies in requirements.txt. - Improved documentation in README.md. - Minor code or logic updates in run.py for enhanced stability or usage.
v1.0.0
Media_Processor 1.2.0 - 新增支持流媒体协议(如 m3u8)及在线视频 URL 自动下载处理。 - 默认支持视频编码从 H.265 (HEVC) 转为 H.264 (AVC)。 - 支持图片和视频的压缩、文字水印及格式转换自动化处理。 - 优化参数说明,丰富输入场景与用法示例。
元数据
Slug media-processor-agent
版本 1.1.0
许可证 MIT-0
累计安装 0
当前安装数 0
历史版本数 2
常见问题

media-processor 是什么?

OpenClaw 核心素材加工工具。支持对图片和视频(含 URL、m3u8、mp4)进行自动化处理,包括下载、压缩、添加水印和格式转换。 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 90 次。

如何安装 media-processor?

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

media-processor 是免费的吗?

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

media-processor 支持哪些平台?

media-processor 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 media-processor?

由 mobai-6543(@mobai-6543)开发并维护,当前版本 v1.1.0。

💬 留言讨论