Creative Space
Edit Your Footage in DaVinci Resolve with an AI Agent

Conteúdo

Provide: The assets to be cut or Resolve project, the purpose of the film and the content you want to express, as well as the duration, aspect ratio and sound requirements. If there are reference films, scripts, subtitles or editing opinions, please provide them together; the confirmed shots, sound and picture processing should indicate the parts that need to be retained.

Receive: A playable video, and a Resolve project corresponding to the finished video that can continue to adjust shot order, cut points and audio tracks. Attached are the assets that actually depend on it and a brief modification description; partial modifications will be delivered within this scope.

The key is to organize the assets into clear expressions: when to give information, when to make it clear to the audience, how the shots connect, and how the sound catches the picture. Choose the cutting method according to the content, and don’t cut all the films into the same fast-paced template.

Usage

  • Cut assets to finished video (default): Select available clips from the assets provided by the user, organize the beginning, main body and end, complete editing and basic sound continuity.
  • Modify existing projects: Adjust shot, rhythm, aspect ratio, subtitles or sound according to specific time points and opinions, retain the remaining confirmed content; work in a new timeline or version.

Tools and responsibilities

  • Host Agent: Read assets, understand modification opinions, organize editing plans and check files; general file and media analysis capabilities are inherited from the host. When operating the interface, read the Computer Use guidance of the current environment and do not use fixed coordinates or a certain menu position as a universal method.
  • DaVinci Resolve (required): Responsible for the actual timeline, basic sound processing, project saving and export. Through the host's Computer Use operation, both the free version and Studio inherit the currently available interface functions.
  • Flova (on demand): Add clearly missing images, illustrations, voiceover, music or sound effects. Follow the Flova Skill to create and obtain the actual files before importing them; users only require to use existing content directly when cutting existing assets.
  • Teardown — retrieve and read the full “Break Down a Reference Video into an Interactive Report” inspiration through Flova MCP by name / storyboard (on demand): When there is a reference film, only the structure, rhythm, cutting points and audio-visual relationships that need to be learned will be analyzed. The existing analysis will be used directly; use storyboard when you need to reshoot, generate shots or reorganize the narrative. When there are clear editing opinions, they will be implemented directly into the corresponding time period.
  • Captions (optional): If requested, offer a suitable HyperFrames, Remotion or FFmpeg post-production workflow.

Confirm the current interface before operation

First confirm that the software can be opened normally, assets can be accessed, and the current project and target timeline can be found.

In the operation interface, first confirm the page, current timeline, selected object and focus. The same shortcut keys in the assets pool, source monitor, and timeline may have different effects. Check the timeline entries and total duration after inserting, appending, or deleting. Reread the interface after a pop-up window, page switch or window change; when the read times out, restore the focus and confirm the current status before deciding whether to retry to avoid repeated insertion.

Make judgment

Determine the order of information first, and then determine the cutting point

Distinguish between the information that must be retained, the images that prove it, and the assets that are just responsible for the atmosphere. Prioritize the audience to understand the subject, action and result; when the same information appears repeatedly, judge whether the repetition adds evidence, contrast or rhythm. When assets lack key processes, explain the gap. Don’t pretend to be irrelevant and have a beautiful shot to make it clear.

Long assets are selected according to the available range, and the entire file is not directly treated as a shot. The shot starts from a valid action or information and ends after the effect is completed. It eliminates camera position finding, mistaken shots, meaningless shaking, and parts where the subject accidentally leaves the frame. Check the facial composition of the character in close-up; explain the environment and actions first, and then go into local details. Clip duration is determined by readable information and actions and is not evenly distributed to total duration.

Spoken delivery uses meaning, breath and tone to organize deletions and avoid cutting off the end of words, changing the original meaning or continuous jump cuts. Action segments select shear points that capture direction, eyeline, action phase, and spatial relationships. Screen recordings and product demonstrations reserve readable time for clicks, interface feedback, and key text. Musical beats can assist organization, so narrative and action don’t have to switch beat by beat.

Transitions are used to explain changes in time, space, or state. First check whether the straight cut can be established, and then choose blending, matching editing or other necessary techniques; for transitions that require assets margin, check the handle length first.

Sound and picture are organized together

First determine the main sound responsible for the information, usually the production audio or voiceover, and then arrange the music, environmental sounds and sound effects. When deleting the picture, check the audio link synchronously to avoid unintentional shifting; let the paragraphs connect naturally by entering the audio early or delaying the exit. Use short fades and necessary ambient undertones to process cut points without masking staccato as excessive noise reduction.

Music gives way to speech, and action sound effects are aligned with actual events. Listen back in full to check for utterances, seams, transitions and clipping; loudness and channeling are determined as delivered. When there are only still frames or no actual sound is heard, it does not claim that the audio and video synchronization and mixing check have been completed.

Workflow

1. Take over assets and projects

Confirm the task, purpose, aspect ratio, duration and delivery form; read assets, original projects, references and modification opinions. Check the actual resolution, frame rate, sound and missing media; keep the existing picture processing and project settings when taking over the project, and only adjust the parts needed this time. First determine the timeline frame rate according to the purpose, and then lay out the tiles; when the assets have the same frame rate and there are no other requirements, the source frame rate will be used and will not be changed by the software default value. Blended frame rates were processed against the target timeline and checked for changes in slow motion and sync sound.

Preserve fallback points with a new project, copy the timeline, or save as a project. After importing the assets into the pool, confirm the actual file; after inserting the first paragraph, check that the video and expected audio track appear on the correct timeline, and the player can see the content before continuing to lay out the video.

2. Build the rough cut

Select shots and available intervals according to the expression goal, and give a brief paragraph sequence or editing instructions. Create a rough cut that can be fully understood, and then check for information omissions, duplications, the connection between actions and eyeline, and the duration of the shot. The existing script and confirmation opinions are used as a basis, and changes will be explained when adjustments have a substantial impact.

Locate and set entry and exit points on the source monitor, and insert or append the selection to the target timeline. Source assets may have their own non-zero timecode; the offset, source timecode and timeline timecode within assets are recorded separately, and "file 5th second" cannot be directly regarded as the source monitor 00:00:05:00. After positioning, confirm according to the screen and selection duration, and then perform editing. When it is necessary to insert in the middle, check whether the subsequent clips are postponed; after deleting the ripple, check the audio and video links, total duration and gaps.

Place the voiceover or production audio in the correct position, and then arrange the auxiliary sound. Supplementary assets only target clear gaps, and retain the corresponding relationship between source files, usage intervals and timelines to facilitate subsequent replacement.

3. Adjust cut points and sound

Adjust cut points, pauses, sound connections and necessary transitions section by section. Preserve existing subtitle tracks outside the requested changes. When the aspect ratio changes, check composition, interface text and existing subtitle safe areas and readability; cropping and exporting alone is not sufficient.

Keep sound and existing subtitles editable, and associate modifications with specific segments or time points.

4. Review, save and export

Play the whole sequence to check opening and ending, cut points, black frames, offline media, transitions, sound, synchronization and existing subtitles. Check action and audiovisual continuity through playback, apply feedback on the corresponding timeline, and confirm the delivered version.

Before exporting, confirm the target timeline and rendering range, remove the remaining entry and exit points or set the selection according to this requirement, and check the output frame rate, aspect ratio and sound. Save the project and export the video; check actual file parameters and expected duration, and replay the exported results.

Deliver the corresponding .drp project and brief modification instructions. .drp does not contain source media; when it needs to be transferred or used across machines, collect the required assets or export the project archive containing the media. Go back to the project to check the media links and final timeline, and avoid handing over unfinished versions or projects that rely on temporary paths.

Storyboard and asset handoff

When new shot planning is needed, the host uses this workflow directly; reuse a clear existing shot list or edit brief. Record each S01 segment’s script link, visual/information goal, duration/time, subject/action, camera or element motion, exact text/sound, transition and production method. Link E01 target elements to user material and actual resources; keep REF-E01 source identities separate and mark unavailable assets as pending. Check identity, space, action and sound continuity; retime to actual speech. For marketing, label applicable hook/turn/proof/cta roles and the viewer gain, using grounded demonstrations; teaching and narrative work retain their own structure. For an unclear expression, find a real reference and record what to borrow/replace. No separately named planning skill or file is needed.

Processo criativo

  1. DaVinci Resolve

    DaVinci Resolve

    Edit the footage

    Arrange shots and adjust cut points and sound.

  2. Flova

    Flova

    Generate additional assets

    Create additional visuals, voiceovers, or music needed for the edit.

  3. DaVinci Resolve

    DaVinci Resolve

    Export the finished video

    Export the video and retain the editable DaVinci Resolve project.