Cutwright
Contents
I built Cutwright to give AI agents a usable editing workflow for client video work. It connects source inspection, transcripts, editorial decisions, motion graphics and render review, so an agent can work through an edit without losing track of which footage and words its decisions refer to.
Keeping the edit attached to the words
A source transcript and a finished edit do not share the same clock. Remove a sentence near the beginning and every later moment moves. A graphic planned against the original timestamp can land on the wrong phrase even when both the graphic and the cut look individually reasonable.
I built a mapping from source words into the cut’s output timeline, preserving the identity and timing of each selected word. Visual events can refer to those anchors, and validation checks whether an anchor still belongs to the chosen source and selected passage.
The edit can change without guessing where the words went.
Separate the decision from the render
The system separates reusable client style and production instructions from each video’s source media, cut, visual plan, rendered graphics and review records. A visual idea can be revised without confusing it with the file currently placed in the timeline. The agent can inspect source footage, an individual graphic or the final composite as different surfaces.
The pipeline
Each stage leaves a file the next one can be checked against, so a problem is caught at the cheapest point: a bad anchor fails the lint before anything renders, and an uncovered cut fails render QC before anyone watches. The agent’s own viewing comes last, and when it finds something, only the affected part is redone.
- Transcribe: Words with timings, straight from the source footage.
- Pack: Transcript, client style and production notes, packed for the model.
- Reason: The model chooses what to keep and where a graphic would help.
- EDL: An edit decision list: the selected passages, in order, as word anchors.
- Lint: Does every anchor still point at a selected word? Caught here, not on screen.
- Visual slots: Optional graphics are rendered into their own slots, apart from the cut.
- Render: Cut and graphics become one file.
- Render QC: Frames around each transition are checked for an uncovered cut.
- Self-eval: The agent watches its own output before calling it ready.
- When self-evaluation finds an issue, the fix re-renders only what changed and the checks run again.
The frame between two shots
A transition could uncover the frame at a camera cut and produce a flash. I added boundary checks around opaque transitions so the edit could identify that timing error before confirming the render as ready.
The workflow produced a full-length rendered edit. Alongside the output, it generates review material and records what still needs attention, so the next editorial pass can start with specific moments to inspect.
Cutwright began from Browser Use’s video-use project. I retained useful rendering and transcription foundations while developing the project structure, word mapping, visual-event validation and review machinery around the needs of this editing workflow.