Free tools and skills for editing on your computer. Copy one prompt, give it to your AI, and it sets the tool up for you. You don't need to know code. You do need a paid AI plan.
copy it, paste it, try it
How it works
Pick one. Start with the one that fixes your biggest headache. You can add more later.
Tap Copy for my AI. It copies a short prompt with the tool's address on .A website for keeping your app’s files online. It also keeps the older versions you send there.For example: Send Lunchbox’s files to GitHub. Later, you can open that saved version from another computer.Still unclear? Ask your AI: “Explain GitHub in even plainer English. Use an example from my app.”
Paste it into an AI that can work on your computer. In the Claude desktop app, that's the Code tab. Codex works too. A regular chat can't set things up on your computer.
Say yes when it makes sense. Your AI reads the tool's own instructions, checks what you already have, and asks before it adds anything big.
Try it on your own clip. It shows you one result. Then it tells you how to use it next time.
Before you start
You need a computer. A Mac or a Windows computer. Phones can't run these.
You need a paid AI plan. Claude or ChatGPT, and the part that works on your computer: the Code tab in the Claude desktop app, or Codex in ChatGPT. A regular chat can't install anything.
The $20 plans usually aren't enough. Claude Pro and ChatGPT Plus are fine to try one tool, but video work runs out fast. I always suggest the bigger plan: Claude Max or ChatGPT Pro.
The tools are free. A few need an account with another company. The card tells you.
Leave some room. Voice and video tools can take a few gigabytes of space.
Add only what you'll use. Each one is one more thing on your computer.
Stay safe
Use the link on the card. Copycat projects exist. Every link here goes to the real one.
Never paste a password into a prompt. If a tool needs an ,API stands for Application Programming Interface.A code that tells a service which app is asking to use it. Some keys are public; secret keys must stay private.For example: Lunchbox uses a key to reach another service. Ask your AI which kind it is before sharing it.Still unclear? Ask your AI: “Explain API key in even plainer English. Use an example from my app.” type it yourself where your AI shows you. Not in the chat.
Your videos stay on your computer. These tools work on your own computer. The card tells you if one uses an online service.
Ask why. Your AI should ask before it installs something for your whole computer. If you're not sure, ask it why it needs it.
What open source means
Every tool here is open source. The people who made it share it for free, and anyone can see how it works. Each one lives in a Your project folder and its saved versions. You can keep a copy on your computer and a copy on GitHub.For example: Lunchbox has its own project folder, with its app files, written plans and older saved versions.Still unclear? Ask your AI: “Explain Repository in even plainer English. Use an example from my app.” on GitHub.
Each repo has a license. It says what you can do with the tool. The ones on this page let anyone use them, even for a business.
Clean up your sound
For videos filmed in the car, the kitchen or anywhere loud.
DeepFilterNet
A noise eraser for your voice.
What it does: Takes background noise out of your voice: road noise, a fan, a hum.
Good for: Videos you filmed in the car or a loud room.
Know this: Works best when the recording is mostly your voice.
Free · License: MIT or Apache 2.0
See what gets copied
Please set up DeepFilterNet on my computer. It's free and open source: https://github.com/Rikorose/DeepFilterNet
1. Read its README on GitHub first and follow its install steps.
2. Check what I already have before you install anything.
3. Ask me before you install anything for my whole computer, and tell me how much space it needs.
4. Don't change my other projects.
5. When it works, try it on one of my own files and show me the result.
6. Then tell me in one or two plain sentences how to use it next time.
What it does: Splits the sound of a video into voice and music, so you can keep just your voice.
Good for: When music or the TV was playing while you talked.
Know this: A long video can take a few minutes on a laptop.
Free · License: MIT
See what gets copied
Please set up Demucs on my computer. It's free and open source: https://github.com/adefossez/demucs
1. Read its README on GitHub first and follow its install steps.
2. Check what I already have before you install anything.
3. Ask me before you install anything for my whole computer, and tell me how much space it needs.
4. Don't change my other projects.
5. When it works, try it on one of my own files and show me the result.
6. Then tell me in one or two plain sentences how to use it next time.
Get to a first cut without doing every snip yourself.
auto-editor
An automatic cutter for the quiet parts.
What it does: Cuts out the quiet parts of a video for you, in one step.
Good for: A fast first cut of a talking video.
Know this: Watch the result. It cuts by sound, so it can't tell a good pause from a bad one.
Free · License: Unlicense (free for anything)
See what gets copied
Please set up auto-editor on my computer. It's free and open source: https://github.com/WyattBlue/auto-editor
1. Read its README on GitHub first and follow its install steps.
2. Check what I already have before you install anything.
3. Ask me before you install anything for my whole computer, and tell me how much space it needs.
4. Don't change my other projects.
5. When it works, try it on one of my own files and show me the result.
6. Then tell me in one or two plain sentences how to use it next time.
What it does: Finds every new shot in a long video and splits it into separate clips.
Good for: After a B-roll day, when you filmed one long video of everything.
Know this: It finds cuts and big changes. A slow pan may stay in one clip.
Free · License: BSD 3-Clause
See what gets copied
Please set up PySceneDetect on my computer. It's free and open source: https://github.com/Breakthrough/PySceneDetect
1. Read its README on GitHub first and follow its install steps.
2. Check what I already have before you install anything.
3. Ask me before you install anything for my whole computer, and tell me how much space it needs.
4. Don't change my other projects.
5. When it works, try it on one of my own files and show me the result.
6. Then tell me in one or two plain sentences how to use it next time.
Captions, cut-outs and voices for the parts you don't film.
whisper.cpp
Free captions, made on your own computer.
What it does: Turns what you say into text, with the time of every word. Use it for captions or a written copy of your video.
Good for: Captions without sending your video to any website.
Know this: It downloads a speech model the first time. Bigger models are slower but more exact.
Free · License: MIT
See what gets copied
Please set up whisper.cpp on my computer. It's free and open source: https://github.com/ggml-org/whisper.cpp
1. Read its README on GitHub first and follow its install steps.
2. Check what I already have before you install anything.
3. Ask me before you install anything for my whole computer, and tell me how much space it needs.
4. Don't change my other projects.
5. When it works, try it on one of my own files and show me the result.
6. Then tell me in one or two plain sentences how to use it next time.
What it does: Takes the background off a photo, so you have just the person or the thing.
Good for: Photos of what you sell, thumbnails and stickers: a cake, a candle, you.
Know this: It downloads a cut-out model the first time it runs.
Free · License: MIT
See what gets copied
Please set up rembg on my computer. It's free and open source: https://github.com/danielgatis/rembg
1. Read its README on GitHub first and follow its install steps.
2. Check what I already have before you install anything.
3. Ask me before you install anything for my whole computer, and tell me how much space it needs.
4. Don't change my other projects.
5. When it works, try it on one of my own files and show me the result.
6. Then tell me in one or two plain sentences how to use it next time.
What it does: A free AI voice that reads your words out loud.
Good for: Reels without your face, or a quick practice voiceover.
Know this: If a reel uses an AI voice, say so in the caption. Never make it sound like a real person.
Free · License: Apache 2.0
See what gets copied
Please set up Kokoro on my computer. It's free and open source: https://github.com/hexgrad/kokoro
1. Read its README on GitHub first and follow its install steps.
2. Check what I already have before you install anything.
3. Ask me before you install anything for my whole computer, and tell me how much space it needs.
4. Don't change my other projects.
5. When it works, try it on one of my own files and show me the result.
6. Then tell me in one or two plain sentences how to use it next time.
A skill is a set of instructions your AI keeps, so it knows how to do one job well. Type its name and it gets to work.
brag
A launch video maker for your app or website.
What it does: Makes a short launch video of your app or website, with music, moving pictures and words to post with it.
Good for: Showing off something you built. It works from your app's folder or a web address.
Know this: Start it by typing /brag. It makes the whole video on your computer, so it can take a while.
Free · License: MIT
See what gets copied
Please add the brag skill to my AI on my computer. It's free and open source: https://github.com/latent-spaces/brag
1. Read its README on GitHub first and follow its steps for adding the skill.
2. Ask me if I want it in every folder or only in one project.
3. Don't change my other skills or projects.
4. When it's added, show me what to type to start it, and try it once with me.
5. Then tell me in one or two plain sentences how to use it next time.
What it does: Lets your AI make animated videos: title cards, captions, lists that pop in, moving graphics.
Good for: Words and graphics on top of your clips, in your own colors and fonts.
Know this: Made by HeyGen. It comes with a few skills. The one called media-use can also make a voice, write captions and take backgrounds off.
Free · License: Apache 2.0
See what gets copied
Please add the HyperFrames skill to my AI on my computer. It's free and open source: https://github.com/heygen-com/hyperframes
1. Read its README on GitHub first and follow its steps for adding the skill.
2. Ask me if I want it in every folder or only in one project.
3. Don't change my other skills or projects.
4. When it's added, show me what to type to start it, and try it once with me.
5. Then tell me in one or two plain sentences how to use it next time.
What it does: Edits a folder of raw clips with you: cuts the ums, adds captions and makes the final video.
Good for: Talking videos when you want your AI to do the first full edit.
Know this: Its captions need an ElevenLabs account and an API key. ElevenLabs has free and paid plans, so check theirs first.
Free skill. Needs an ElevenLabs account · License: MIT
See what gets copied
Please add the video-use skill to my AI on my computer. It's free and open source: https://github.com/browser-use/video-use
1. Read its README on GitHub first and follow its steps for adding the skill.
2. Ask me if I want it in every folder or only in one project.
3. Don't change my other skills or projects.
4. When it's added, show me what to type to start it, and try it once with me.
5. Then tell me in one or two plain sentences how to use it next time.
It needs an ElevenLabs API key. When you get to that step, show me where to type it myself. Don't ask me to paste it into this chat.
What it does: A big set of video skills: screen recordings of your app, simple edits, voices and more.
Good for: Recording a demo of your app for a reel.
Know this: More technical. Some parts use paid services. Your AI should tell you which before it sets them up.
Free. Some parts use paid services · License: MIT
See what gets copied
Please add the Claude Code Video Toolkit skill to my AI on my computer. It's free and open source: https://github.com/digitalsamba/claude-code-video-toolkit
1. Read its README on GitHub first and follow its steps for adding the skill.
2. Ask me if I want it in every folder or only in one project.
3. Don't change my other skills or projects.
4. When it's added, show me what to type to start it, and try it once with me.
5. Then tell me in one or two plain sentences how to use it next time.
Only set up the parts I need: screen recording to start. Tell me before you set up anything that uses a paid service.
What it does: Helps your AI write a skill of your own, like your editing style or the way you like captions.
Good for: Once you edit the same way every time and want your AI to remember it.
Know this: Made by Anthropic, the company behind Claude. Its folder has its own license file.
Free · License: Apache 2.0
See what gets copied
Please add the Skill Creator skill to my AI on my computer. It's free and open source: https://github.com/anthropics/skills/tree/main/skills/skill-creator
1. Read its README on GitHub first and follow its steps for adding the skill.
2. Ask me if I want it in every folder or only in one project.
3. Don't change my other skills or projects.
4. When it's added, show me what to type to start it, and try it once with me.
5. Then tell me in one or two plain sentences how to use it next time.