Install Gemini

Install in Gemini

Before you start: Gemini skills need you to be 18 or over, signed in, with Keep Activity turned on. Work and school Google accounts may not have skills yet. See Requirements and Recommendations.

Gemini uses the same build of the skills as ChatGPT and Grok (the chatgpt/ and chatgpt-prompt/ folders). There are two ways in:

• The Gemini app (gemini.google.com or the Gemini app on Mac): you upload the skill folder.
• Gemini CLI on your own computer: one command per skill, straight from GitHub.

Gemini app (web and Mac)

1. Download the zips from the latest release:

   • videoexpress-scripting-chatgpt.zip for VideoExpress Scripting
   • videoexpress-prompt-chatgpt.zip for the companion skill, VideoExpress Prompt (optional: a step-by-step wizard for one image or clip prompt)

   Use the chatgpt zips, not the claude ones: the Claude build is worded for Claude.

2. Unzip each one. Gemini needs SKILL.md at the top of what you upload, and in these zips it sits one folder down. On Windows, right-click the zip → Extract All; on a Mac, double-click it. Inside you'll find a videoexpress-scripting (or videoexpress-prompt) folder holding SKILL.md and a references folder. That inner folder is the one you upload.

3. Open gemini.google.com on your computer, or the Gemini app on a Mac.

4. In the sidebar click Customize (under More if you don't see it), then Skills at the top.

5. Click Create → Upload, select the videoexpress-scripting folder, and click Open.

6. Review the skill and click Create. Repeat steps 5 and 6 with the videoexpress-prompt folder if you want the companion skill.

7. Start a new chat and try "I've got an idea for a VideoExpress video. Help me plan it." Gemini picks the skill up when it's relevant; to call it by name, type / (or @) and choose it.

Uploading works on a computer. Once a skill is created it's also available in the Gemini mobile app.

Updating

The Gemini app doesn't update skills from GitHub. When a new version is released, download and unzip the new zip, then on the Skills page click the skill → More → Replace skill and upload the new folder.

Gemini CLI

In Gemini CLI on your own computer, install each skill straight from GitHub:

gemini skills install https://github.com/ratava/videoexpress-scripting-skill.git --path chatgpt/skills/videoexpress-scripting
gemini skills install https://github.com/ratava/videoexpress-scripting-skill.git --path chatgpt-prompt/skills/videoexpress-prompt

The second line is the companion skill; leave it out if you don't want it. Skills install for your user by default; add --scope workspace to install into the current project only. Check they loaded with /skills list inside a session.

Use these paths, not the repo root. The repo also holds a Claude build of each skill with the same name.

To update, uninstall and install again:

gemini skills uninstall videoexpress-scripting
gemini skills install https://github.com/ratava/videoexpress-scripting-skill.git --path chatgpt/skills/videoexpress-scripting

Skills installed this way are only seen by Gemini CLI on that computer, not by the Gemini app.

Letting Gemini drive VideoExpress

The assisted and automatic operating modes need an assistant that can control a browser. In Gemini CLI that comes from a browser MCP server (for example Chrome DevTools). Where your Gemini can't control a browser, use the manual mode: Gemini plans and writes every prompt, and you paste them into VideoExpress yourself.