Automate with Skills (Platform: Power Your AI Tools with Conductor Data (MCP))
Are you a Conductor customer?
Review this lesson—and the rest of our educational content and activities—in our in-platform Community!
Not a customer? Click here to view this lesson
By now you know how to query your data conversationally and turn insights into deliverables in the moment. Skills take that workflow and package it — so instead of guiding the AI step by step each time, you hand it a playbook and it runs the sequence for you.
What is a Skill?
Here's the honest truth: the users who get the best, most consistent results from Conductor's MCP won't be the ones who query most — they'll be the ones who run Skills.
Imagine asking your AI to run a complete audit of your brand's AI search visibility — covering brand mentions, citation share, sentiment, competitive positioning, and a prioritized action plan — and having it done, structured, and formatted in under two minutes. That's what a Skill does. Not because the AI suddenly got smarter, but because someone already did the hard work of engineering the workflow.
Technically, a Skill is a plain-text markdown file called SKILL.md. It contains a set of structured instructions that tells your AI assistant exactly what data to pull, in what order, how to analyze it, and how to present the output. You load it once, say "run it," and the AI executes the whole sequence — no step-by-step guidance, no re-explaining what you want, no formatting the output yourself afterward.
The SKILL.md format is an open standard built on agentskills.io — created specifically to make AI workflows portable and reusable across different tools and platforms. The idea: you write a workflow once, and it works wherever AI works. Claude and ChatGPT have natively adopted the format. As the standard spreads, Skills you build or download today will work in more and more places over time — without modification.
The first time you do a brand visibility audit conversationally, you're essentially teaching the AI as you go — asking questions, following up, nudging the output into shape. A Skill skips all of that. It's the playbook already written, tested, and ready to run. The difference isn't small: what takes a 20-minute back-and-forth becomes a 90-second command.
Conductor's Skills template gallery is built by people who know this data deeply — the endpoints, the metrics that matter, what good AEO analysis actually looks like. So you're not starting from a generic prompt and hoping it works. You're starting from a workflow that's already been engineered to get you to a useful answer fast.
The experience of loading and running a Skill varies by platform — more on that in a moment.
Conductor's Skill Template Gallery
Conductor has a growing library of pre-built Skills designed for the most common AEO workflows. A few highlights...
Audit Your Brand's AI Search Visibility
The flagship skill. Run a complete audit of your brand's visibility, citations, and sentiment across AI engines, with a prioritized action plan at the end. The kind of analysis that used to take hours, ready in minutes.
Build a Board-Ready AI Search Deck
Take your AI search performance data and walk out with a clear, defensible story formatted for an executive audience. From raw data to presentation-ready output in one command.
Read the AI Search Pulse Verdict
Not every workflow needs to be a full audit. This one gives you a 30-second verdict: is your brand winning or losing in AI search this week? Fast, sharp, and perfect for a Monday morning check-in.
The full, growing library is at conductor.com/docs/templates/gallery. Filter by your persona (SEO/AEO, Content, Executive, Brand/PR, E-Commerce, and more) or by use case (competitive, reporting, content, brand research) to find what's most relevant to your role. You can download skills individually or grab the whole set at once.
Claude and ChatGPT — fully supported, recommended
Both platforms have native Skill-loading features built for the SKILL.md format. This is where you'll get the cleanest, most reliable experience.
In Claude:
Go to Settings → Features → Skills, click the + button, and upload the skill's ZIP file.
Once loaded, the skill appears in your Skills menu and can be invoked in any conversation.
In ChatGPT:
Go to Skills, click New Skill → Upload from your computer, and select the skill file.
Once loaded, it's available to use in your chats.
Other platforms — try it, results may vary
Since Skills are plain markdown files, they work on any AI platform that can process text instructions. If you're using Perplexity or another tool without a native Skills feature, you can paste the contents of a SKILL.md file at the start of a conversation or upload it as a reference document. Many AI assistants handle the instructions well — just know you'll need to paste or upload each session, and results will vary more than with native support.
Microsoft Copilot — a note before you try
We know many of you live in Copilot, and we want to be straight with you: Skills are not reliably compatible with M365 Copilot right now, and we'd hate for you to hit a wall and wonder what you did wrong.
The issues aren't with the skill files. They're with how M365 Copilot handles long markdown instruction sets, combined with enterprise security policies that vary by organization and can interfere with custom instruction processing. In practice, this means Skills may execute partially, behave inconsistently, or return errors that have nothing to do with anything you did.
Copilot works great for conversational querying — the How to Query Your Data and Prompt Starter Pack sections are fully applicable to Copilot users. Use it for that. For running structured Skill workflows, we'd recommend pulling up Claude or ChatGPT — both are available in many enterprise environments alongside Copilot, and the Skills experience there is exactly what it's designed to be.
SKILL.md is becoming an industry-wide standard, and we're hopeful that as adoption grows, Microsoft will bring native Skills support to M365 Copilot. When that happens, your full Conductor toolkit will work wherever you work. We'll update this course when that changes.
Make it yours
Conductor's Skills deliver real, structured output right out of the box. But every team's data, competitors, and priorities are different, and the real compounding value comes from tuning a skill to your context.
After running a skill, you might find you want to:
Adjust the output format to match your team's reporting template
Filter by a particular persona or change the time period
Add a section specific to how your team measures success
Do that in the conversation, the same way you'd refine any AI response: just ask. Once you've shaped the output the way you want it, ask the AI to make it permanent:
Update my skill with the changes we've made — I want this version to run by default going forward.
In Claude and ChatGPT, the AI regenerates an updated SKILL.md with your preferences. At that point, you're not running the Conductor "Audit Your Brand's AI Search Visibility" skill anymore. You're running your company's version of it, with your competitors, your format, and your priorities baked in. Every run gets faster from there.
The first run teaches you what you want. Every run after that arrives faster and more tailored.