What you can do
Choose a task, prepare the right material and understand the result to expect.
Check where each capability is available
This guide describes rikoko MCP and its local workspace. The web app has a separate interface and may expose fewer operations. Local editing works with existing assets; AI analysis and generation require configured model services. Ask your connected agent which features are available before starting. Model support does not guarantee visual fidelity or a successful generation.
Edit existing footage
Provide a video or accessible local file and the change you want. Trim, split, reorder or replace footage; adjust volume and aspect ratio; add existing music or captions. The result is an edited export and a workspace link. These local operations do not buy new generated media. Changing the frame fit does not imply automatic person tracking.
Trim my video to its first 5 seconds, keep the original sound and export an MP4. Use existing footage only.Analyze a reference video
Provide a reference video and say whether you want a summary, a shot breakdown or a detailed adaptation plan. A configured analysis model can return observations about shots, action, text, sound and pacing. Analysis does not create a new video and may incur model charges. Unclear details should remain marked as uncertain.
Analyze this reference shot by shot. Identify its opening, actions and pacing. Show me any analysis cost first; do not generate a new video.Create or edit an image
Provide a description and, when needed, product, character or scene images. With a compatible image model, create an image, edit a reference or prepare the first frame of a video. The output is a still image; animating it is a separate video task. Image preparation is optional and can have its own cost.
Use my product photo to plan an opening image on a kitchen table. Preserve the product identity. Show the image plan and cost before generating.Generate or adapt a video
Start with a brief, an opening image or supported subject and motion references. State what should change and what must stay. A compatible video profile returns generated clips that can be assembled into your edit. Reference modes, duration and resolution vary by profile. A final 15-second edit may use several shorter clips. Adapting a reference does not guarantee a frame-by-frame match.
Plan a vertical product video using my product images and the pacing of this reference. Explain what can be preserved, which model supports the inputs and the proposed cost before generating.Add or revise captions
Provide a video with speech, an audio file or existing timed captions. A configured transcription service can supply text and word timing; editing existing captions can reuse the current visuals. Review spelling and timing before export. Transcribing audio and changing a caption’s appearance are separate operations; automatic translation quality is not guaranteed.
Change the existing caption at 00:03 to “Available now”. Keep the visuals, timing and sound unchanged. Export a new version.Create voiceover, music or sound effects
For voiceover, provide the script, language and voice choice. Voice cloning additionally needs an authorized voice recording accepted by the model; save the resulting voice profile for reuse. Music and sound effects use a description and supported duration. These services produce audio assets; specify where to place them and how to balance them with the original sound.
Plan a voiceover from my script in my chosen language. Keep the video unchanged and lower the background music under speech. Show the voice option and cost before generating audio.Export and continue editing
After rendering, receive a video attachment or playable media link plus a local workspace link. The client determines how media is displayed. Continue in the same conversation or open the workspace to inspect and edit. Saved revisions let you return to earlier work. Re-exporting a completed file does not require buying new generation; a new edit can require another local render.