📌 Gemini Skills: Build Once, Reuse Anywhere, Combine Multiple Ones (Full Guide in 9 Mins)

Introduction

Google announced Gemini Skills as the eventual replacement for Gems. As of early October, Skills are already live in Gemini Spark and rolling out in Google Workspace.

The rollout into the Gemini app itself starts October 13, so plenty of personal accounts won't see Skills for a few more days.

So if you open the Gemini app today and still see Gems, that's fine. The app rollout hasn't reached every account yet, and it's gradual, with completion expected by mid-November.

Your existing Gems keep working until migration, and they'll convert to Skills automatically when the time comes.

In this guide, I’ll show you how to create Gemini Skills, use them for repeated tasks, and combine multiple Skills when your workflow becomes more complex.

I. What are Gemini Skills?

Gemini Skills are reusable workflows you save once and call up whenever you need them. Instead of pasting the same long instructions into Gemini over and over, you build the Skill once and invoke it with a / slash command in any chat.

Gemini can also apply an enabled Skill automatically when it's clearly relevant.

A Skill can include:

• Custom instructions: what Gemini should do and how

• Context: background information it needs to handle the task properly

• Output rules: what the result should look like

• Supporting files: PDFs, CSVs, style guides, or reference documents

You can also combine multiple Skills in the same chat when a task has several steps. More on that below.

How Skills Differ From Gems

Skills are still catching up to Gems on a few features. As of October 2026, Skills don't yet support Canvas, Deep Research, Create video, Create music, or Guided Learning.

If your workflow depends on any of those, test before you fully switch over. Google says more of the old Gem functionality, along with Google Drive file links and Skill sharing, is coming over the next few weeks.

II. How to Create Gemini Skills

There are 3 main ways to create Gemini Skills. If the workflow is simple and clear in your head, the direct prompt method is fastest.

If you've already worked through a task with Gemini and refined it over several messages, turn that conversation into a Skill instead.

All three entry points live in Settings → Skills on the Gemini web app, Mac app, or mobile.

Key points

• Create Skills from prompts, SKILL.md, or past conversations.

• Define clear input, process, output, and rules.

• Add supporting files when the workflow needs real data.

Method 1: Create a Gemini Skill With a Prompt

If you already know the task you want to repeat, describe the role, input, process, output, and rules in one prompt. The more specific you are upfront, the more consistent the Skill's output will be later.

Here's an example for a Meeting Summary Skill:

Create a reusable Gemini Skill called "Meeting Summary."

Purpose:
Turn raw meeting notes or transcripts into a clear summary that I can review or share.

Input:
I may provide a full transcript, rough notes, or both.

Process:
- Identify the main topics discussed.
- Separate confirmed decisions from ideas or suggestions.
- Extract action items only when there is a clear task.
- Add the owner and deadline only when they appear in the source.
- Merge repeated points and remove small talk.
- Do not guess missing names, dates, or decisions.

Output:
- Meeting Summary
- Key Decisions
- Action Items
- Open Questions

Rules:
- Keep the summary concise but complete.
- Use bullet points when listing decisions, actions, or questions.
- Keep important names, dates, numbers, and commitments.
- Mark anything uncertain instead of making assumptions.

Save this as a reusable Gemini Skill and keep these rules every time I use it.

Gemini Skills work better when you give Gemini the full job upfront.

Tell it what kind of input to expect, how you want it handled, what the output should look like, and where it shouldn't make assumptions. Vague Skills produce inconsistent results.

Method 2: Upload a Skill File

If you already have a SKILL.md file, for example one you've built for Claude or another platform, you can upload it directly instead of rebuilding the workflow from scratch.

Go to Settings → Skills, then click the Upload icon. From there, add your SKILL.md file (or a folder or .zip with SKILL.md in its main folder) plus any supporting files the Skill needs, such as CSVs, PDFs, pricing sheets, or style guides.

Important Note: Gemini Skills use the same open SKILL.md format as Anthropic's Agent Skills, so a well-structured Skill you built for Claude imports directly. You don't have to rebuild from zero. Two things to know:

• import currently works only on the Gemini web app and the Mac app, and the total size of all files in one Skill can't exceed 100 MB.

• The Skill name inside SKILL.md also has to be lowercase with words separated by hyphens, like meeting-summary.

Once you upload the files, Gemini reads them and builds the Skill from that content. After you review it, the new Skill appears under Active on the Skills page.

Example use case: a Proposal Writer Skill could use SKILL.md for the instructions and attach a CSV file with real pricing data, so Gemini uses your actual prices instead of guessing.

Method 3: Turn a Conversation Into a Skill

This is my favorite method for complex workflows, cause the process has already been tested in a real conversation. You're saving what already worked.

Use this once you've refined a workflow over several messages with Gemini:

Review everything we did in this conversation and turn the final workflow into a reusable Gemini Skill.

Keep only the steps, rules, decisions, and output format that helped produce the final result.

Remove failed attempts, repeated instructions, temporary examples, and anything we later changed.

Structure the Skill with:
- Purpose
- Required input
- Step-by-step process
- Output format
- Rules and constraints
- Error handling

Don't add new rules that were not established in this conversation.

Before saving the Skill, show me the final structure so I can review it.

It’s pretty the same as Claude Skills right.

III. How to Use Gemini Skills in a Chat

Once you've created a Skill, using it is simple. In any Gemini chat, type / and choose the Skill you want from the list that appears.

That's it. Then add your new input. The instructions and output format are already saved inside the Skill, so you don't rewrite them.

For example, if I call the Meeting Summary Skill, I only need to paste the transcript or notes. Gemini already knows what to do with it and what the output should look like.

My usual flow:

1. Type / and select the Skill

2. Add the new input for this specific task

3. Check the output before using it

For repeated tasks, this is the part that saves the most time. You skip the long prompt, Gemini follows the rules you already set, and the output stays consistent across different inputs.

IV. How to Combine Multiple Gemini Skills

When one task has several steps, you can use multiple Gemini Skills in the same workflow instead of putting every rule inside one Skill.

Here, I’ll use Proposal Writer Skill with Proposal Reviewer Skill.

• Proposal Writer Skill creates the proposal from the client brief and supporting files

• Proposal Reviewer Skill checks pricing, scope, unsupported claims, and missing details before you send it

You can call both Skills in Gemini Chat and use this prompt:

Use both the Proposal Writer Skill and Proposal Reviewer Skill for this task.

First, use Proposal Writer to create the proposal from the client brief and supporting files I provide.

Then, use Proposal Reviewer to review the completed proposal.

The review must check:
- Pricing and calculations
- Scope and deliverables
- Unsupported claims or guarantees
- Missing or unclear information
- Any detail that needs human confirmation

If the reviewer finds a real issue, revise the proposal and run the review again.

Return:
1. Final Proposal
2. Review Summary
3. Final Status: Ready to Send or Needs Human Review

Do not invent missing facts, prices, deadlines, discounts, or promises.

If the supporting files do not confirm something, mark it clearly instead of guessing.

This setup keeps each Skill focused on one job. One drafts, the other checks, and you get a cleaner, more trustworthy output without building a monster prompt that tries to do both at once.

V. How to Manage Your Gemini Skills Library

Once you've created a few Skills, go to Settings → Skills to manage them.

Click the three-dot menu next to any Skill to:

• Use it right away

• Edit it with Gemini's help

• Edit it manually

• Download it

• Replace it with an updated version

• Deactivate it

• Delete it

If a Skill uses supporting files and those files need updating, say a new pricing sheet, use Replace skill and upload the new files.

You can create as many Skills as you want, but only 100 can be active at once. If you build one for every tiny task, finding the right one becomes its own problem. Focus on Skills for tasks that genuinely repeat.

VI. The Best Ways to Use Gemini Skills

Gemini Skills are worth building for tasks that repeat often, follow a clear process, and benefit from consistent output. They're not worth creating for tasks that change every time or that you'll only do once.

For me, the best Gemini Skills are the ones I can reuse without thinking too much about setup.

Skip building a Skill when the task changes significantly each time, you'll only do it once or twice, or the instructions are short enough to paste quickly anyway.

Conclusion

Gemini Skills start making sense the moment you stop treating every task like a brand-new conversation.

Once you save the right instructions, supporting files, and output rules, repeated work gets much easier. You can reuse a Gemini workflow, combine multiple Gemini Skills when the task gets bigger, and keep the process consistent without rebuilding it every time.

The best place to start is simple: pick one task you already repeat, turn it into a Gemini Skill, and test it on real work. If it saves you from explaining the same thing to Gemini again tomorrow, it’s already doing its job.

If you are interested in other topics and how AI is transforming different aspects of our lives or even in making money using AI with more detailed, step-by-step guidance, you can find our other articles here:

• GPT Images 2.5 is Seriously Impressive: Every New Feature (King of AI Images?)

• I Tested GPT-6.1 Sol Inside Codex on 3 Hard Tasks. Can THIS Beat Claude Sonnet 5.5?

• 5 Free AI Agent Skills Tested for Research, Coding, Video, Security and Better UI*

• Sonnet 5.5 is Here! And It Can Make Seriously Good Videos (5 Incredible Use Cases)*

• The Ultimate AI Setup for Windows: Turn It Into an AI Super-Machine (Full Guide)*

  *indicates a premium content, if any