Capcut Edit logo

Capcut Edit

Community
renezander030
capcut-edit

Edit CapCut / JianYing video projects — read and write subtitles, timing, speed, volume, templates, animations (fade/ken-burns), and cut long-form to shorts. Use when the user mentions capcut, jianying, subtitles, video editing, draft_content.json, draft_info.json, or cutting videos — in English or Chinese (剪映, 字幕, 草稿, 剪辑, 切片, 视频编辑).

Overview

Publisherrenezander030
Repositorycapcut-cli
Skill namecapcut-edit
Stars
871
Forks
90
Bundled files
13
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 13 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by renezander030 on GitHub. Read the source before you install it.

Installation

Install the Capcut Edit AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/renezander030/capcut-cli/tree/master/skills/capcut-edit
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/renezander030/capcut-cli.git /tmp/capcut-cli
mkdir -p .claude/skills
cp -r /tmp/capcut-cli/skills/capcut-edit .claude/skills/capcut-edit
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Capcut Edit in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Capcut Edit on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Capcut Edit is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

capcut-edit

CLI for editing CapCut / JianYing drafts through a version-aware store. v0.11 detects and synchronizes readable timelines across draft_content.json, draft_info.json, draft_meta_info.json, and template-2.tmp.

Project locations

  • macOS: ~/Movies/CapCut/User Data/Projects/com.lveditor.draft/<project>/
  • Windows: %LOCALAPPDATA%\CapCut\User Data\Projects\com.lveditor.draft\<project>\

Core principle — deterministic scripts

Any recipe involving more than one capcut call, any arithmetic, or any branching ships as a parameterised shell script in scripts/. Claude invokes the script with inputs; the script runs a fixed sequence. If a recipe can be a script, it must be a script — not a narrative in references/.

references/workflows.md documents which script to call and why, never a reconstructable sequence.

Progressive disclosure

Start broad, drill into what you need. Never dump full project JSON.

bash
capcut info <project> -H           # overview
capcut doctor -H                   # installed media/transcription capabilities
capcut diagnose <project> -H       # canonical files, divergence, editor safety
capcut tracks <project> -H         # all tracks
capcut materials <project> -H      # material summary
capcut segments <project> -H       # segments with timing
capcut segment <project> <id>      # one item, full detail

Output modes

  • JSON (default): pipe to jq, feed to scripts.
  • -H: human-readable tables.
  • -q: quiet, exit code only — for write commands in scripts.

Batch writes

bash
echo '{"cmd":"set-text","id":"a1b2c3","text":"Fixed"}
{"cmd":"volume","id":"d4e5f6","volume":0.5}' | capcut batch <project>

Operations: set-text, shift, shift-all, speed, volume, opacity, trim.

Batch is transactional: one failing operation writes nothing. Use --continue-on-error only when the user explicitly accepts a partial commit; it still exits non-zero.

Where to look

  • references/api-reference.md — every command, every flag, every value format (time, percentages, degrees), the ID-prefix rule, the .bak invariant.
  • ../../docs/command-reference.md — generated command-contract v2 with typed positionals/options, mutability, prerequisites, output form, and exit codes.
  • references/workflows.md — recipes = which scripts/X.sh to run and why.
  • references/pitfalls.md — close-project-first, .bak, clip=null on audio, source_timerange math with speed, alpha-keyframes-don't-render-use-animation-materials, etc.
  • Enum discovery — call capcut enums --<category> (or -H for a table) to list every CapCut slug. Categories: --transitions, --masks, --image-intros, --image-outros, --image-combos, --text-intros, --text-outros, --text-loop-anims, --scene-effects, --character-effects, --audio-effects, --fonts. Add --jianying to switch namespace. The enums.json bundle is generated once via python3 scripts/extract-enums.py; runtime is Python-free.
  • scripts/ — the client library. Current:
    • anim.sh — attach any of 9 CapCut intro/outro animations; fade-in.sh and fade-out.sh are thin wrappers.
    • ken-burns.sh — scale + pan keyframes via capcut keyframe --batch (motion properties render; alpha does not — use anim.sh fade-in instead).
    • long-to-short.sh — cut a range, stamp title + CTA text.
    • stamp-cta.sh — apply a saved text template (see assets/examples/subscribe-cta.json).
    • _test.sh — run every wrapper against the fixture; run after any change here.
  • assets/examples/ — raw JSON snippets for hand-editing.

Invariants

  • Close the project in CapCut before editing; reopen after.
  • Every write is conflict-checked, prepared atomically, and synchronized to every readable timeline target. Every target receives .bak and rolling history snapshots.
  • Never pass --force-write unless the user explicitly accepts overwriting changed-on-disk state or an open editor.
  • clip is null on audio segments (no opacity/scale).
  • capcut cut writes to --out, never modifies the source.
  • IDs: first 6+ chars of the UUID match as a prefix.
  • Time formats: 1.5s, 500ms, +0.5s, -1s, 1:30, 0:05.5, bare number = seconds.

v0.11 high-value paths

  • Whole draft: capcut compile spec.json --check, then capcut compile spec.json --out <dir>. Specs support refs, transforms, decorators, templates, captions, keyframes, and fades.
  • Karaoke: capcut caption <project> --audio <path> --whisper-engine <openai|whisper-cpp|faster-whisper> --karaoke.
  • Proxy: capcut render <project> --all-video-tracks --burn-captions; run doctor first for FFmpeg capability flags.
  • Automation: capcut serve --workers N --retries N --timeout MS; give jobs stable id fields and let the runner serialize writes per project.
  • Agent schema: on CLI v0.28.0 or newer, start with capcut describe --compact to discover commands, then fetch complete contracts with capcut describe --command <name> (repeat for several names). Plain capcut describe returns the full registry and works on older versions. Use these contracts as the source of truth; do not scrape help text or invent flags.

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Capcut Edit AI skill do?

Edit CapCut / JianYing video projects — read and write subtitles, timing, speed, volume, templates, animations (fade/ken-burns), and cut long-form to shorts. Use when the user mentions capcut, jianying, subtitles, video editing, draft_content.json, draft_info.json, or cutting videos — in English or Chinese (剪映, 字幕, 草稿, 剪辑, 切片, 视频编辑).

Why use Capcut Edit on TypingMind?

Because you install it once and use it with any model. Capcut Edit is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Capcut Edit in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/renezander030/capcut-cli/tree/master/skills/capcut-edit. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Capcut Edit?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Capcut Edit?

As many as you like. As long as a model supports skills, you can use Capcut Edit with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Capcut Edit AI skill free?

Yes. It is published on GitHub by renezander030 under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇