Usage Rules
- First read
../byted-mediakit-shared/SKILL.md, and perform the unified pre-check; if this Skill is missing, stop and prompt for installation. - Select only from the following table
editingdomain tools; similar capabilities are distinguished by each tool's "capability description" and parameter boundaries. - Before execution, read the corresponding reference as needed; parameter and result descriptions come from the same reviewed copy, and the complete machine contract is subject to the current CLI
--schema. - When required parameters, authentication environment variables, or real input resources are missing, ask the user for them; generic optional fields may only be passed through with values explicitly provided by the user, other optional fields may be accurately determined from clear intent, but must not be fabricated.
- During execution, set
MEDIAKIT_SURFACE=skill, specifying that the invocation source is Skill. - During execution, set
MEDIAKIT_RUNTIMEto the current Agent host, to avoid the CLI being unable to reliably identify the parent Agent.
Clarification and Cross-Domain Routing
If only the media type is given without specifying whether to trim, merge, 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 image; video intelligent processing such as video quality enhancement, content understanding, subtitle extraction from video, speech-to-subtitle, subtitle removal, invisible watermarking, portrait or green screen matting should be routed to video; non-editing objectives such as audio transcoding, voice endpoint detection, and vocal-background separation should be routed to audio. Filter effects include spring/dayglow/bright/fair/food, etc.; specific parameters are subject to the tool reference.
Tool List
| Tool | Description | Supported Modes | Command | Reference |
|---|---|---|---|---|
| add-image-to-video | Supports overlaying a specified image (such as a Logo or watermark) onto the video frame. | Cloud / Local | mediakit-cli editing add-image-to-video |
reference/add-image-to-video.md |
| add-subtitle-to-video | Burns a subtitle file or text content onto the video frame in a custom style, generating a new video with embedded subtitles. | Cloud / Local | mediakit-cli editing add-subtitle-to-video |
reference/add-subtitle-to-video.md |
| adjust-audio-speed | Used for audio speed adjustment, allowing adjustment of audio playback speed to achieve fast or slow playback effects. | Cloud / Local | mediakit-cli editing adjust-audio-speed |
reference/adjust-audio-speed.md |
| adjust-video-speed | Used for video speed adjustment, producing fast or slow playback effects by adjusting the playback speed. | Cloud / Local | mediakit-cli editing adjust-video-speed |
reference/adjust-video-speed.md |
| adjust-video-volume | Used to adjust the volume level of the input video, and can also achieve muting. | Cloud / Local | mediakit-cli editing adjust-video-volume |
reference/adjust-video-volume.md |
| apply-camera-motion | Adds a camera movement effect to the input video within a specified time period, commonly used in scenarios such as secondary creation of materials, marketing openings, and short drama motion effects. | Cloud | mediakit-cli editing apply-camera-motion |
reference/apply-camera-motion.md |
| apply-video-filter | Adds a filter effect to the specified video. | Cloud | mediakit-cli editing apply-video-filter |
reference/apply-video-filter.md |
| concat-audio | Concatenates multiple audio clips. | Cloud / Local | mediakit-cli editing concat-audio |
reference/concat-audio.md |
| concat-video | Concatenates multiple videos in order into a complete video file, and supports adding transition effects at the splice points. | Cloud / Local | mediakit-cli editing concat-video |
reference/concat-video.md |
| crop-video | Crops the video frame according to a specified rectangular region, and the crop result retains only the specified required region. | Cloud | mediakit-cli editing crop-video |
reference/crop-video.md |
| extract-animated-image | Captures a segment of footage from a video according to a specified start time and end time, generating a GIF or WebP animated image, commonly used for making cover animations, marketing materials, and short previews. | Cloud | mediakit-cli editing extract-animated-image |
reference/extract-animated-image.md |
| extract-audio | Separates the audio track from the input video file, generating an independent audio file. | Cloud / Local | mediakit-cli editing extract-audio |
reference/extract-audio.md |
| fade-audio | Apply a fade-in or fade-out effect to the start and end positions of the input audio, and output the processed audio file. | Cloud / Local | mediakit-cli editing fade-audio |
reference/fade-audio.md |
| fade-video-audio | Perform fade-in or fade-out processing on the input video's audio track at the beginning or end, used to soften abrupt starts and stops of the audio track and improve the listening experience of the final video. Output the processed video file. | Cloud / Local | mediakit-cli editing fade-video-audio |
reference/fade-video-audio.md |
| flip-video | Used for video image flipping, mirroring the specified video vertically or horizontally. | Cloud / Local | mediakit-cli editing flip-video |
reference/flip-video.md |
| image-to-video | Combine multiple images in order into a dynamic video, with configurable transition animations and in-shot animations; only turns existing images into a video with motion effects, and does not support generating new visual content based on reference images. | Cloud | mediakit-cli editing image-to-video |
reference/image-to-video.md |
| mix-audio | Mix multiple audio files (such as background music, sound effects, and vocals) to generate a new audio file. 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 duration) is 1. The duration of the output audio is based on the longest audio. Output video format: mp3 | Cloud / Local | mediakit-cli editing mix-audio |
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 video's original audio track; when the audio and video durations are inconsistent, alignment processing can be performed. | Cloud / Local | mediakit-cli editing mux-audio-video |
reference/mux-audio-video.md |
| rotate-video | Used for video image rotation, rotating the specified video as a whole. | Cloud | mediakit-cli editing rotate-video |
reference/rotate-video.md |
| stitch-video | Stitch multiple videos spatially horizontally or vertically into a single complete frame, suitable for scenarios such as multi-view comparison and frame composition. | Cloud | mediakit-cli editing stitch-video |
reference/stitch-video.md |
| text-to-scrolling-video | Convert the specified text content into a scrolling text video. The output video is fixed at 9:16 vertical format, commonly used for scenarios such as novel promotions, content explanations, and lyric videos. | Cloud | mediakit-cli editing text-to-scrolling-video |
reference/text-to-scrolling-video.md |
| trim-audio | Used for audio trimming, extracting a segment from the input audio according to the specified start time and end time. | Cloud / Local | mediakit-cli editing trim-audio |
reference/trim-audio.md |
| trim-video | Used for video trimming, extracting a segment from the input video according to the specified start and end times. | Cloud / Local | mediakit-cli editing trim-video |
reference/trim-video.md |