ACCESSIBILITY STATEMENT

Google Workspace Just Got Skills: How to Use Them

  • by Sandy Waggett
  • 11 min reading time
Cream and deep teal editorial graphic with the headline Turn Your Rules Into a Reusable AI Skill, three numbered setup cards, a laptop showing a checklist, and MSW Interactive Designs branding

Google Workspace now supports reusable Skills. Start with a simple proposal quality check so Gemini follows your pricing and scope rules, with you in control.

TL;DR: Google announced Skills in Google Workspace on November 10, 2026. Skills are reusable instructions plus reference files that tell Gemini how to do a defined job your way. Start small. Build a Proposal Quality Check Skill that compares a draft to your approved pricing and scope, returns a clear review list, and leaves final decisions to you.

If you want AI to help, give it your rules and a tight lane. Skills in Workspace let you package a repeatable task so Gemini applies your pricing, templates, and checklists instead of winging it.

Do not roll this out across your company on day one. Pick one job that is narrow, testable, and low risk. A proposal quality check fits that bill. The Skill reviews, you decide. No sending, no editing the original.

What changed and why you should care

Google’s Workspace Updates post, “Scale customized AI across your team with skills in Google Workspace and Workspace Studio,” confirmed Skills are rolling out in Gmail, Docs, Slides, Drive, and Chat. A Skill is a saved set of instructions with optional reference files that guide Gemini to handle a repeatable job using your team’s rules. You can create, test, and share Skills in Google Workspace Studio and in Google Docs. You can also create or edit Skills in context inside Gmail, Docs, Slides, Drive, and Chat.

Once Skills are turned on for your domain, Gemini can automatically apply a relevant Skill to your content, or you can invoke a Skill directly with an @mention. Skills can also trigger Workspace Studio flows using a “Start manually” or “On a schedule” starter step. Skills are not yet available in Google Sheets or Google Meet, and they currently support English prompts and instructions only.

The business case is simple. AI gets more useful when you give it approved knowledge, a clear job, and boundaries. A Skill packages all three. You spend less time retyping instructions, and you reduce risk by encoding your non-negotiables.

Choose one job: The Proposal Machine, scoped to a review

Here is the example we use with clients: The Proposal Machine. The agent turns sales-call notes, approved services, and pricing into a proposal draft for human review. That full workflow is powerful, but you do not need to start there. Start with one piece, the part with the least risk and the highest clarity: a quality check.

Your first Skill will not draft or send anything. It will review a proposal draft against your pricing and scope checklist, then return a short list of findings for you to evaluate. This is a focused, repeatable job you can test in under an hour.

Prepare your inputs: what to gather and how to structure it

You are telling Gemini how your business writes and prices proposals. Put the right files in place before you build the Skill:

  • Pricing and terms: your current price list, fees, discounts you actually offer, and standard payment or delivery terms you approve.
  • Approved service descriptions: short scope language for each service you sell. These are the canonical blurbs the team should use.
  • Proposal checklist: the items every proposal must include (client identifiers, scope boundaries, inclusions and exclusions, assumptions, timelines, signature block, and required disclosures).
  • A redacted sample proposal: remove or mask client names, addresses, or any regulated data. Deliberately omit one required item so you can see if the Skill flags it.

Store these references in Google Drive in a clearly named folder. Use stable filenames and keep a separate, offline backup so you are not relying on a single copy inside your workspace. When pricing or services change, plan to update the files and the Skill instructions on the same day.

Build the Proposal Quality Check Skill in Workspace Studio

Here is the step-by-step path. This keeps you in control and gives you a testable result.

  1. Open Google Workspace Studio. If you are new to the tool, review the Workspace Studio help guide so you know where Skills live and how sharing works.
  2. Create a new Skill. Name it “Proposal Quality Check.” Write a one-line description so teammates know exactly what it does. Example: “Reviews proposals against our pricing and scope checklist, returns a findings list.”
  3. Add reference files. Attach your current pricing, service descriptions, and proposal checklist from Drive. Keep them small and specific. The Skill does not need your entire sales folder.
  4. Paste the instructions below into the Skill’s instruction field. These are written to review only. No editing, no approvals, no sending.
Review this proposal against my attached pricing and scope checklist. Flag missing deliverables, pricing discrepancies, unclear scope, and unsupported claims. For each issue, quote the proposal wording and identify the relevant reference. If information is missing or sources disagree, ask me. Never invent prices, terms, or guarantees. Return a short review list. Do not edit the original, approve the proposal, or send anything.
  1. Save and test in Docs. Open your redacted sample proposal in Google Docs, then ask Gemini to apply @Proposal Quality Check. Confirm it returns a concise list of issues, with quotes and references, and that it did not edit the original.
  2. Probe the boundaries. Ask it to approve or send the proposal. It should decline. Change a price in the draft and rerun the Skill to see if it flags the discrepancy. Your goal is to confirm the guardrails hold.
  3. Document the playbook. Capture where the Skill lives, which references it uses, how to invoke it (@mention and when auto-application makes sense), and when to escalate findings to a human owner.

When you are satisfied, you can share the Skill from Workspace Studio with your sales or account team. Sharing creates an independent copy for the recipient. If the Skill references Drive files, you must grant access to those files separately. Keep a master copy in your team folder and a private offline copy so you can roll back if needed.

Run, review, and govern: how to use the Skill day to day

Make it boring and reliable. That is the point.

  • Invocation. Gemini can auto-apply relevant Skills, or your team can @mention the Skill in Docs, Gmail, or Chat when they need it. For proposals, I recommend an explicit @mention so the human stays in the loop.
  • Output handling. The Skill should return a short, numbered review list with proposal quotes and references. The human proposal owner decides what to change and when to send. No automatic edits.
  • Studio flows. If you want a light workflow wrapper, connect the Skill to a Workspace Studio flow with a “Start manually” starter step. Example: a teammate clicks to run the check, then the flow posts the findings to a shared Chat space for visibility. If you use “On a schedule,” keep it to internal review documents, not client drafts.
  • Version control. When prices or services change, update the Drive files first, then update the Skill instructions if the rules changed. Note the effective date in the Skill description.
  • Audit and training. Spot check one proposal per week. If the Skill misses something, fix the instruction or the reference file. If a teammate ignores findings, address the human, not the AI.

What this will not do: it will not guarantee proposal accuracy, revenue growth, or close rates. It reduces rework and catches obvious misses. People remain accountable for what gets sent.

Availability, prerequisites, and honest boundaries

Here are the constraints you need to know before you plan a rollout:

  • Editions. Skills are available for Business Starter, Standard, and Plus; Enterprise Starter, Standard, and Plus; Google AI Pro for Education; and AI Expanded Access. Users must be 18 or older.
  • Rollout. Rapid Release domains: gradual rollout starting November 9, 2026, up to 15 days for feature visibility. Scheduled Release domains: starting November 30, 2026, also up to 15 days.
  • Defaults and admin controls. Skills are ON by default for eligible editions. An admin can disable them at the domain, organizational unit, or group level.
  • Where they work. You can create, test, and share Skills in Google Workspace Studio and Google Docs. You can create or edit Skills in context in Gmail, Docs, Slides, Drive, and Chat. Skills are not yet available in Sheets or Meet. English prompts and instructions only at this time.
  • Sharing model. When you share a Skill from Workspace Studio, the recipient gets an independent copy. If a Skill references Drive files, share those files separately or the Skill will not be able to cite them.
  • Gemini app distinction. Skills in Workspace are separate from skills and Gems in the Gemini app and do not sync automatically. Both use the open SKILL.md format, but treat them as different surfaces.
  • Data handling. Keep a copy of your instructions and reference files off-platform. If you lose access, you should still have your rules. When pricing or services change, update both the off-platform copy and the Workspace version the same day.

Prerequisites for success are simple: a clear job, accurate references, one owner to maintain them, and a short test loop. If any of those are missing, fix that first.

Not on Workspace? Build the Proposal Machine in another tool

If you do not use Google Workspace or your edition is not eligible yet, you can still build a small agent to draft and review proposals. Claude Cowork, Manus, or ChatGPT Work can all support a reusable workflow with attached references. Claude Code is another path if you prefer a more technical setup. Features, permissions, and access vary by platform.

Start with the same references you prepared above: your current pricing, approved service descriptions, proposal checklist, plus your last three approved proposals and redacted sales-call notes. Then use this build prompt as your starter specification. Paste it into your chosen platform and follow its setup steps. Keep the same boundaries: no editing source files and no sending anything automatically.

Build prompt:
Help me build a reusable Proposal Machine workflow. Review my approved proposals, current pricing, service descriptions, scope checklist, and sales-call notes. Ask about gaps or conflicts first. Use past proposals for structure, not current prices or client facts.

Draft from the approved inputs, then check against my pricing and scope rules. Return the draft, a separate review list with source references, and unresolved questions. Never invent facts, prices, discounts, deadlines, or guarantees. Do not edit source files or send anything. Stop for my approval.

Save the approved instructions in this tool's supported reusable format. Give me a downloadable copy and steps to run it again. Identify any setup still needed. Do not enable automatic runs.

Whichever tool you choose, require a human approval step. If the platform supports scheduling, leave it off for proposal work. Proposals affect pricing and commitments. Keep people in the loop.

If you prefer structured, done-with-you support to set up repeatable marketing and sales workflows, our MSW Performance Marketing Program focuses on practical systems you can run consistently.

Common pitfalls and how to avoid them

  • Vague instructions. If you say “improve this,” you will get creative rewrites. If you say “return a findings list with quotes and references,” you will get a review you can act on. Be explicit.
  • Outdated references. Old prices create new headaches. Assign a single owner to update the price list and Skill when rates change. Note the effective date.
  • Missing access. Your team can have the Skill and still be blocked by Drive permissions. Share files intentionally and test as a typical user, not just as admin.
  • Over-scoping. Do not try to encode your entire sales process on day one. Start with the review. Expand later if the review proves reliable.
  • Skipping tests. Always test with a redacted sample and a seeded error. Confirm the Skill catches what it should and refuses what it should not do.

Your next step

Pick one recent proposal, gather your pricing, approved service descriptions, and checklist, then build and test the Proposal Quality Check Skill. Keep an offline copy of the instructions and reference files, and update both when your pricing or services change. AI drafts and checks. You decide what gets sent.

For ongoing ideas like this, our running list lives in the MSW AI Tips archive.

About the author: Sandy Waggett is the Founder of MSW Interactive Designs and InfiniteYous. She helps businesses implement practical AI tools and connected marketing systems that support real-world visibility, content, websites, and operations.

Tags


Marketing Strategies That Work Blog

  • Cream and deep teal editorial graphic with the headline Turn Your Rules Into a Reusable AI Skill, three numbered setup cards, a laptop showing a checklist, and MSW Interactive Designs branding

    by Sandy Waggett Google Workspace Just Got Skills: How to Use Them

    Read more 

  • Sandy Waggett beside the headline Claim Your Business in ChatGPT, with find, claim and verify steps and MSW branding

    by Sandy Waggett ChatGPT Business Profiles Are Here: How to Claim Your Business

    Read more 

  • MSW editorial cover asking where your site is missing in AI search, with cited-source cards and a highlighted content gap

    by Sandy Waggett Bing's Topics Portfolio Shows Where Your Site Is Missing

    Read more 

  • MSW cover showing the same buyer question across ChatGPT, Meta AI and Gemini, under the headline AI Visibility Goes Beyond ChatGPT

    by Sandy Waggett AI Visibility Beyond ChatGPT: Why Meta AI and Gemini Count

    Read more 

Book a Strategy Call

It's free of charge but insanely valuable. We'll use the time the way you want to. Website or marketing review, AI integration ideas, you name it. Let's start a conversation - we're here to help you be successful online!


Footer image

© 2026 MSW Interactive Designs LLC, Powered by Shopify

    • American Express
    • Apple Pay
    • Diners Club
    • Discover
    • Google Pay
    • Mastercard
    • Satispay
    • Shop Pay
    • Visa

    Login

    Forgot your password?

    Don't have an account yet?
    Create account