Communitygithub.com

edit

Builds a video version from the user's recording: the speaker cut, then one edit plan with all inserts (2D scenes, 3D scenes, recordings, keywords), sound and music, built with one command (crewcut render). Use for: edit the video, build the cut, speaker cut, rough cut, insert, b-roll, 2D scene, 3D insert, counter, keyword, music bed, sound, assemble, render the preview, new plan version, short from the long video.

What is edit?

edit is a Claude Code agent skill that builds a video version from the user's recording: the speaker cut, then one edit plan with all inserts (2D scenes, 3D scenes, recordings, keywords), sound and music, built with one command (crewcut render). Use for: edit the video, build the cut, speaker cut, rough cut, insert, b-roll, 2D scene, 3D insert, counter, keyword, music bed, sound, assemble, render the preview, new plan version, short from the long video.

Works with~Claude Code~Codex CLI~Cursor
npx skills add https://github.com/marcopatzelt/crewcut/tree/HEAD/skills/edit

Ask in your favorite AI

Open a new chat with this agent skill pre-loaded.

Documentation

Edit: one plan, one command

The Editor turns the user's recording into a video in two parts. First the speaker cut: the best takes, cut at pauses, graded. Then the edit plan (edit/plan-vNNN.json): every insert, its words and cues, the sound. The workbench crewcut render builds the plan. Nothing in this skill holds a colour, a font or a level: those come from the style pack (crewcut brand show). Taste lives in brand/rules/edit.md and brand/rules/cut.md.

Read first

You want toRead
cut the recording into the speaker cutreferences/record-to-cut.md (sound to the end, split files, last full take, double starts, cuts at pauses, re-shoot, grade, show and lock)
write the plan, choose or build a scene, show a recordingreferences/plan-and-scenes.md (word anchors, count the mix, form per statement, cues, new 2D scene, reuse, recordings)
make music, set levels, or rent a GPUreferences/sound-and-gpu.md (music along the story, chosen by measurement, levels, GPU pitfalls, the cost card)
build a 3D sceneskill 3d-scene
check before showingskill cut-check (run by the Critic, never by you)

Scene names and their data keys: templates/2d/BLOCKS.md and templates/3d/README.md. Do not copy key lists here.

How a version is made

  1. Speaker cut (references/record-to-cut.md): crewcut ingest, crewcut transcribe, crewcut cut …. Run python3 skills/cut-check/scripts/cut_check.py <decisions> on it and fix every ERROR before you build on it.
  2. Words of the final cut: crewcut transcribe audio/voice-vNNN.wav → work/words-vNNN.json. All plan times come from these words.
  3. Count the mix, then list the spots (section below). Send the list as a card before the big render: crewcut card propose --episode $CREWCUT_EPISODE --step edit --kind mix --title "Inserts for v002" --file edit/mix-v002.md. Keep building while it is open; the card is a look, not a gate.
  4. Write the plan with a small generator script (edit/make_plan.py) that sets every time from a word anchor. Never type seconds by hand. Schema: docs/architecture.md, "Edit plan". Values left null come from the pack.
  5. crewcut render plan edit/plan-vNNN.json: lists every insert with frames and cues. Read it. An insert of half a second is a wrong anchor.
  6. crewcut render 2d edit/plan-vNNN.json: 2D inserts, locally.
  7. 3D: crewcut render probe edit/plan-vNNN.json <id> gives one still frame at half size. Look at it next to the reference before anything else (skill 3d-scene). Then the cost estimate and the card (section "Rented GPU or local"), then crewcut render 3d edit/plan-vNNN.json --rented --gpus N or --local.
  8. crewcut render cut edit/plan-vNNN.json: assembly with the music bed, hits, typing, gaps, ducking and the loudness target from the pack. The output is unchecked. Tell the Showrunner: "built, unchecked, do not show".
  9. Hand the version to the Critic (crewcut job add --episode $CREWCUT_EPISODE --identity critic --step edit "check plan-vNNN"). Only after both checks show 0 ERROR: the preview goes to export/, the Showrunner shows it as a card with the file (crewcut card propose … --kind preview --file export/<id>-vNNN-preview.mp4), and right after showing: crewcut render shown edit/plan-vNNN.json. From then on every change is a new plan version, even for two frames.

A changed text or number: change the plan (new version), run steps 5 to 9. Only changed inserts render again. crewcut render all <plan> runs 2d, 3d and cut in one go when nothing needs a look in between.

Mix first, then the plan

These hold for every plan, before one frame is rendered. The limits come from edit-style.json.

  1. Symbol before text. Every insert first shows something that happens. A text card (headline, post) only where the viewer must read a real source: at most text_cards_max sources and at most text_share_max of the full-screen inserts. keyword only for a number or a name, at most keywords_max, each at most keyword_max_words words. Count by kind before you render; the style check counts again.
  2. Symbolic means 2D and 3D. 3D where an object with an action is the idea. Drawn 2D for processes, amounts, ways and comparisons: shapes that build, move or change. Per spot, pick what a viewer understands in 2 seconds. A scene holds at most one word or one number, on the spoken word.
  3. The bar is the reference, not "it renders". Compare with baseline.json and brand/references/. Rented GPUs are for quality: full resolution, enough samples, soft shadows, clean edges. 2D is motion design, not a slide: parts build one after the other, the camera drifts a little, transitions are soft.
  4. Check every scene before the cut. A still frame and a strip of 5 frames next to the reference. If it looks like a slide deck, clip art or a template, throw it away and build it again. Few very good scenes beat many average ones; reuse a scene 2 to 4 times with another word or state.
  5. Real material, without being asked. A post, an announcement or a third-party result goes into the picture: the post as a card with a mark on the spoken word, a third-party video short on a card with its source visible. Never an invented screenshot, never a company logo you made up.
  6. Music belongs to the story (references/sound-and-gpu.md). Instruments only, quiet under the voice, with a source and rights note.
  7. Compute fast, never idle. While machines load, keep working locally (plan, 2D, cut). Every remote step has a time limit. At the end prove that no machine runs.
  8. Show the list before the big render (step 3): time, kind, one line of what one sees, the number of text cards.

Scenes

2D scenes (HTML, rendered by Chromium, styled by the brand tokens) and 3D scenes (Blender by code, templates/3d/). Data keys: templates/2d/BLOCKS.md, templates/3d/README.md. Which form for which statement: references/plan-and-scenes.md, section 3.

KindScenes
2D numbers and amountscounter, bars, grid, wave
2D sources and real materialheadline, post, clip-card, cards, swap
2D drawn ideasarea, flow, untangle
2D over the speaker (transparent)keyword
3Dobject-word, row, grid3d, stack, slide, box, scale (shared kit templates/3d/stage.py)
recording"kind": "clip" with file, at, speed: a finished recording full screen

The neutral pack ships few, general 3D scenes. A user's own scenes go to brand/templates/ (never edit the app's templates). A new scene gets a line in BLOCKS.md or templates/3d/README.md and is not tested until a frame of it has been looked at.

Mix rules (values from the pack)

  • 3D share: about graphics_mix.3d of the inserts; 2D carries the numbers; the speaker is on screen in between.
  • Cut on the word, not on the beat. The word in the picture lands 0 to 3 frames before it is spoken.
  • Split long speaker stretches with "speaker": [{"at": s, "zoom": z}] (zoom from zoom_base to zoom_punch, never over zoom_max).
  • The speaker is never shorter than speaker_min_seconds between two full-screen parts, and at least speaker_open_close_seconds at the start and the end.
  • Sound: music bed at music_db with duck, a hit only at strong spots, typing under counters, a short gap in the music before the most important number. Loudness loudness_lufs, true peak at most true_peak_db.
  • Sources honest: a headline card names the real source.
  • Never claim a viewing or listening check that did not happen. This session cannot hear: name the times.

When you apply a rule that comes from a named source, you may say so in one line, for example crewcut say --episode $CREWCUT_EPISODE --cite "fast cutting and keyword captions after Alex Hormozi's editing style" "…".

Rented GPU or local

  • 3D frames render in Blender Cycles. Local means the container's CPU: fine for probes at half size and very short scenes. Anything longer goes to a rented GPU.
  • Cost first. crewcut compute estimate --kind blender --frames N [--local-sec-per-frame S] prints time and money for local and rented. Put the estimate on a card (--kind spend, title with the amount), wait with crewcut card wait <id>. Exit 3 means no: render locally or drop the scene. Only an identity with may_spend: true rents (the Editor).
  • Always: still frames at half size first, then the animation.
  • Only scene code and fonts go to a rented machine. Never footage, the voice, keys or private data.
  • At the end: crewcut compute destroy --all-ours, then crewcut compute ledger --episode ID; the report to the user holds the spend and the proof that no machine runs. Also after a failure.
  • Pitfalls and what the tools do about them: references/sound-and-gpu.md, section B.

Honest limits

  • The agent cannot hear. Every music and voice judgement is a listening list with times for the user.
  • Detection of a word rest, a double start or a level jump finds candidates, not certainties.
  • The style check measures limits; whether a scene looks good only a person can say. Put the frames next to the reference and say plainly which scenes are still below it.

Planned

  • Re-timing 3D renders when word times shift by a few frames, so they are not rendered again.
  • Generated music on a rented GPU through crewcut compute (the studio did it by hand; see references/sound-and-gpu.md).
  • Web page recordings as a render step (Chromium is in the container; no command yet).
  • Comparison sheets (scene frame next to the reference) as one command.
  • A self-test on the demo episode in examples/demo/ once it exists.

Sources

No connection to or endorsement by these people; we took ideas and wrote our own summary.

Individual skills in this repo

This repo contains 3 individual skills — each has its own dedicated page.

script

Writes the spoken script of an episode in the user's own voice (voice.md), in teleprompter form, after title and proof are fixed: opening that delivers the title's promise, small stories with rising stakes, the critical side, a verdict in its own form, the channel's closing line. skills/script/scripts/script_check.py checks form, sentences, opening, ending, numbers against the proof, and length. Delivered through the web teleprompter (crewcut prompter). Use when: \"write the script\", \"teleprompter text\", \"rewrite the script\", \"shorten the script\", \"make it sound natural\", \"script for the episode\". Not for titles or thumbnails. Short scripts: the `--short` mode of the check and skills/shorts.

shorts

Makes YouTube Shorts in two modes: Clip (cut from a finished episode: candidates from the transcript and retention, a 9:16 rough cut from the original recording around the face, portrait graphics, music with clear rights, the Critic's own Short check) and Original (a Short recorded for itself: three ideas with a proof from the user's real work, the user talks freely, the Writer shortens, only the speaker is recorded, the crew brings the B-roll). Publishing: one card per Short, or the user's own standing approval for Shorts; at most two a day. Use when: \"make shorts\", \"make a short about …\", \"clip this episode\", \"short from episode N\", \"publish shorts without asking\", \"how are my shorts doing\". Not for long videos (skills/episode, skills/edit).

title

Writes and checks the YouTube title and the thumbnail text of an episode, before the script. Delivers five candidates with a hook and a reason each, one recommendation, and three variants for the A/B test; skills/title/scripts/title_check.py checks every candidate against the style pack. Use when: \"title\", \"video title\", \"YouTube title\", \"what should the video be called\", \"check this title\", \"shorten the title\", \"better title\", \"thumbnail text\". Not for the blog's search title and not for the thumbnail image.

Related Skills