byted-mediakit-editing
Compare original and translation side by side
🇺🇸
Original
English🇨🇳
Translation
Chineseediting MediaKit Skill
editing MediaKit Skill
使用规则
Usage Rules
- 先读取 ，执行统一前置检查；该 Skill 缺失时停止并提示安装。
../byted-mediakit-shared/SKILL.md - 只从下表选择 域工具；相似能力按各工具“能力描述”和参数边界区分。
editing - 执行前按需读取对应 reference；参数与结果说明来自同一份已审核文案，完整机器合同以当前 CLI 为准。
--schema - 缺少必填参数、鉴权环境变量或真实输入资源时，向用户索取；通用可选字段只能透传用户明确提供的值，其他可选字段可由明确意图准确确定，但不得伪造。
- 执行时设置 ，指定调用来源是 Skill。
MEDIAKIT_SURFACE=skill - 执行时把 设置为当前 Agent 宿主，避免 CLI 无法可靠识别父级 Agent。
MEDIAKIT_RUNTIME
- First read and perform unified pre-checks; stop and prompt for installation if this Skill is missing.
../byted-mediakit-shared/SKILL.md - Only select tools in the domain from the table below; distinguish similar capabilities based on each tool's "capability description" and parameter boundaries.
editing - Read the corresponding reference as needed before execution; parameter and result descriptions come from the same reviewed document, and the complete machine contract is subject to the current CLI .
--schema - Request information from users when required parameters, authentication environment variables or real input resources are missing; only pass through values explicitly provided by users for general optional fields, other optional fields can be accurately determined based on clear intent, but cannot be forged.
- Set during execution to specify that the call source is a Skill.
MEDIAKIT_SURFACE=skill - Set to the current Agent host during execution to avoid the CLI failing to reliably identify the parent Agent.
MEDIAKIT_RUNTIME
澄清与跨域路由
Clarification and Cross-Domain Routing
若只给出媒体类型而未说明要剪、合、叠、调、加滤镜还是分析，应先澄清。纯图像增强、抠图、OCR、图片水印或图片尺寸优化应路由到 image；视频画质增强、内容理解、从视频提取字幕、语音转字幕、字幕擦除、暗水印、人像或绿幕抠像等视频智能处理应路由到 video；音频转码、语音端点检测、人声背景分离等非剪辑目标应路由到 audio。滤镜效果包括春日/晚霞/鲜亮/白皙/食物等，具体参数以工具 reference 为准。
If only the media type is provided without specifying whether to cut, combine, overlay, adjust, add filters or analyze, clarification should be made first. Pure image enhancement, matting, OCR, image watermarking or image size optimization should be routed to the image domain; video intelligent processing such as video quality enhancement, content understanding, subtitle extraction from video, speech-to-subtitle, subtitle erasure, invisible watermarking, portrait or green screen matting should be routed to the video domain; non-editing targets such as audio transcoding, speech endpoint detection, vocal and background separation should be routed to the audio domain. Filter effects include spring/sunset/vibrant/fair/food, etc., and specific parameters are subject to tool references.
工具列表
Tool List
| 工具 | 说明 | 支持模式 | 命令 | 参考 |
|---|---|---|---|---|
| add-image-to-video | 支持将指定图片（如 Logo、水印等）叠加到视频画面上。 | Cloud / Local | | reference/add-image-to-video.md |
| add-subtitle-to-video | 将字幕文件或文本内容按自定义样式压制到视频画面中，生成带内嵌字幕的新视频。 | Cloud / Local | | reference/add-subtitle-to-video.md |
| adjust-audio-speed | 用于音频调速，可调整音频播放倍速，实现快放或慢放效果。 | Cloud / Local | | reference/adjust-audio-speed.md |
| adjust-video-speed | 用于视频调速，通过调整播放倍速产生快放或慢放效果。 | Cloud / Local | | reference/adjust-video-speed.md |
| adjust-video-volume | 用于调整输入视频的音量大小，也可实现静音。 | Cloud / Local | | reference/adjust-video-volume.md |
| apply-camera-motion | 对输入视频在指定时间段内添加一种运镜特效，常用于素材二次创作、营销片头、短剧动效等场景。 | Cloud | | reference/apply-camera-motion.md |
| apply-video-filter | 为指定视频添加滤镜效果。 | Cloud | | reference/apply-video-filter.md |
| concat-audio | 拼接多个音频片段。 | Cloud / Local | | reference/concat-audio.md |
| concat-video | 将多个视频按顺序拼接成一个完整的视频文件，并支持在拼接处添加转场效果。 | Cloud / Local | | reference/concat-video.md |
| crop-video | 按指定的矩形区域裁剪视频画面，裁剪结果仅保留指定的需要区域。 | Cloud | | reference/crop-video.md |
| extract-animated-image | 从视频中按指定开始时间和结束时间截取一段画面，生成 GIF 或 WebP 动图，常用于制作封面动图、营销素材和短预览。 | Cloud | | reference/extract-animated-image.md |
| extract-audio | 从输入视频文件中分离音轨，生成独立的音频文件。 | Cloud / Local | | reference/extract-audio.md |
| fade-audio | 对输入音频的起止位置实现淡入或淡出效果，输出处理后的音频文件。 | Cloud / Local | | reference/fade-audio.md |
| fade-video-audio | 在片头或片尾对输入视频音轨执行淡入或淡出处理，用于弱化音轨突兀的起止，提升成片听感。输出处理后的视频文件。 | Cloud / Local | | reference/fade-video-audio.md |
| flip-video | 用于视频画面翻转，对指定视频进行上下或左右镜像翻转。 | Cloud / Local | | reference/flip-video.md |
| image-to-video | 将多张图片按顺序组合成动态视频，可配置转场动画和镜头内动画；仅把现有图片做成带动效的视频，不支持根据参考图生成新的画面内容。 | Cloud | | reference/image-to-video.md |
| mix-audio | 将多个音频文件（如背景音乐、音效、人声）进行混音，生成一个新的音频文件。<br>处理耗时：处理耗时与视频时长正相关。视频时长越长，处理耗时越长。平均 RTF（处理耗时/原片时长）为 1。<br>输出音频的时长以最长的音频为准。<br>输出视频格式：mp3 | Cloud / Local | | reference/mix-audio.md |
| mux-audio-video | 可将输入的音频流与视频流合并成一个新的视频文件，并可选择保留或替换视频的原有音轨；当音视频时长不一致时，可进行对齐处理。 | Cloud / Local | | reference/mux-audio-video.md |
| rotate-video | 用于视频画面旋转，对指定视频进行整体旋转。 | Cloud | | reference/rotate-video.md |
| stitch-video | 将多个视频在空间上按水平或垂直方向拼接成一个完整画面，适用于多视角对比、画面组合等场景。 | Cloud | | reference/stitch-video.md |
| text-to-scrolling-video | 将指定文本内容转换为文字滚屏视频，输出视频为固定 9:16 竖版，常用于小说推文、内容讲解和歌词视频等场景。 | Cloud | | reference/text-to-scrolling-video.md |
| trim-audio | 用于音频裁剪，按指定的开始时间和结束时间从输入音频中截取片段。 | Cloud / Local | | reference/trim-audio.md |
| trim-video | 用于视频裁剪，可按指定的开始和结束时间从输入视频截取片段。 | Cloud / Local | | reference/trim-video.md |
| Tool | Description | Supported Mode | Command | Reference |
|---|---|---|---|---|
| add-image-to-video | Supports overlaying specified images (such as logos, watermarks, etc.) onto video frames. | Cloud / Local | | reference/add-image-to-video.md |
| add-subtitle-to-video | Embeds subtitle files or text content into video frames with custom styles to generate a new video with embedded subtitles. | Cloud / Local | | reference/add-subtitle-to-video.md |
| adjust-audio-speed | Used for audio speed adjustment, which can change the playback speed of audio to achieve fast or slow playback effects. | Cloud / Local | | reference/adjust-audio-speed.md |
| adjust-video-speed | Used for video speed adjustment, which produces fast or slow playback effects by changing the playback speed. | Cloud / Local | | reference/adjust-video-speed.md |
| adjust-video-volume | Used to adjust the volume of the input video, and can also achieve mute effect. | Cloud / Local | | reference/adjust-video-volume.md |
| apply-camera-motion | Adds a camera motion effect to the input video within a specified time period, commonly used in scenarios such as material re-creation, marketing openings, and short drama animations. | Cloud | | reference/apply-camera-motion.md |
| apply-video-filter | Adds filter effects to the specified video. | Cloud | | reference/apply-video-filter.md |
| concat-audio | Splices multiple audio clips together. | Cloud / Local | | reference/concat-audio.md |
| concat-video | Splices multiple videos in sequence into a complete video file, and supports adding transition effects at the splicing points. | Cloud / Local | | reference/concat-video.md |
| crop-video | Crops the video frame according to the specified rectangular area, and only retains the specified required area in the cropped result. | Cloud | | reference/crop-video.md |
| extract-animated-image | Captures a segment of frames from the video according to the specified start and end time to generate GIF or WebP animated images, commonly used for making cover animations, marketing materials and short previews. | Cloud | | reference/extract-animated-image.md |
| extract-audio | Separates the audio track from the input video file to generate an independent audio file. | Cloud / Local | | reference/extract-audio.md |
| fade-audio | Implements fade-in or fade-out effects at the start and end positions of the input audio, and outputs the processed audio file. | Cloud / Local | | reference/fade-audio.md |
| fade-video-audio | Performs fade-in or fade-out processing on the audio track of the input video at the opening or ending to weaken the abrupt start and end of the audio track and improve the listening experience of the finished video. Outputs the processed video file. | Cloud / Local | | reference/fade-video-audio.md |
| flip-video | Used for video frame flipping, which performs vertical or horizontal mirror flipping on the specified video. | Cloud / Local | | reference/flip-video.md |
| image-to-video | Combines multiple images in sequence into a dynamic video, and supports configuring transition animations and in-lens animations; only converts existing images into videos with motion effects, does not support generating new frame content based on reference images. | Cloud | | reference/image-to-video.md |
| mix-audio | Mixes multiple audio files (such as background music, sound effects, human voices) to generate a new audio file.<br>Processing time: Processing time is positively correlated with video duration. The longer the video duration, the longer the processing time. The average RTF (processing time/original video duration) is 1.<br>The duration of the output audio is based on the longest audio.<br>Output audio format: mp3 | Cloud / Local | | reference/mix-audio.md |
| mux-audio-video | Can merge the input audio stream and video stream into a new video file, and can choose to retain or replace the original audio track of the video; when the audio and video durations are inconsistent, alignment processing can be performed. | Cloud / Local | | reference/mux-audio-video.md |
| rotate-video | Used for video frame rotation, which performs overall rotation on the specified video. | Cloud | | reference/rotate-video.md |
| stitch-video | Splices multiple videos horizontally or vertically in space into a complete frame, suitable for scenarios such as multi-angle comparison and frame composition. | Cloud | | reference/stitch-video.md |
| text-to-scrolling-video | Converts specified text content into a text scrolling video, the output video is fixed 9:16 vertical format, commonly used in scenarios such as novel promotion, content explanation and lyric videos. | Cloud | | reference/text-to-scrolling-video.md |
| trim-audio | Used for audio cropping, which captures a segment from the input audio according to the specified start and end time. | Cloud / Local | | reference/trim-audio.md |
| trim-video | Used for video cropping, which captures a segment from the input video according to the specified start and end time. | Cloud / Local | | reference/trim-video.md |