menu
I Built Reusable Marketing Skills for GitHub Copilot — Here's How (and Why)

I Built Reusable Marketing Skills for GitHub Copilot — Here's How (and Why)

posted Originally published at dev.to 4 min read

What if your code editor could do keyword research, audit your SEO, and optimize your content for AI search engines — without leaving VS Code?

I built a set of open-source agent skills that turn GitHub Copilot into a hands-on marketing strategist. Here's what I learned, how they work, and how you can use (or build) your own.

The problem
I run a small website. Every time I wanted to write a blog post, the workflow looked like this:

Open Ahrefs/Semrush → research keywords
Open Google Docs → write the article
Open WordPress → publish and fill in SEO fields
Open ChatGPT → rewrite for social media
Repeat
Five tools. Five context switches. And every time, I found myself giving the same SEO instructions to AI assistants: "Use the keyword in the first paragraph", "Write a meta description under 155 characters", "Structure with H2/H3 headings"...

What if I could encode all that knowledge once, and have my AI assistant apply it automatically?

The solution: Copilot Skills
GitHub Copilot (in agent mode) can read SKILL.md files from your workspace. These files act as knowledge modules — they tell Copilot how to approach a domain, what rules to follow, and what frameworks to apply.

I created four skills that cover the full marketing workflow:

seo-keyword-research → Find what your audience searches for

    ↓

seo-strategy → Structure your site to rank

    ↓

geo-optimization → Get cited by AI search engines

    ↓

content-creator → Write content that ranks AND gets cited
What's inside a skill?
A skill is just a Markdown file with structured knowledge. Here's a simplified look at the GEO (Generative Engine Optimization) skill:

name: geo-optimization
description: "Optimize content for AI search engines like ChatGPT,
Gemini, and Perplexity. Use when the user mentions 'AI visibility',

'GEO', 'AI search', or 'cited by AI'."

GEO — Generative Engine Optimization

What is GEO?

GEO = optimizing content so AI-powered search engines cite your
brand in generated answers.

GEO Scorecard (16 criteria)

Criterion Target
Authoritative tone Expert language, no hedging
Structured data Schema markup present
...

Anti-patterns

Don't do this Why
Keyword stuffing AI recognizes and skips it
...

The description field is key — Copilot uses it to decide when to invoke the skill. The body contains the actual knowledge the AI applies.

How it works in practice
Here's what a real session looks like:

Me: "Do keyword research for a blog about home espresso machines"

Copilot (automatically invokes seo-keyword-research):

Searches the web for related keywords
Classifies intent (informational, commercial, transactional)
Analyzes SERP features
Outputs a structured keyword map with primary/secondary keywords, difficulty estimates, and content type recommendations
Me: "Now write the article and optimize it for AI search"

Copilot (invokes content-creator + geo-optimization):

Writes with proper H2/H3 structure
Includes the keyword in the first paragraph
Adds statistics with citations (GEO requirement)
Uses authoritative tone
Ends with a clear conclusion (AI search engines need this)
Generates meta title and description within character limits
Me: "Publish as draft to WordPress"

Copilot (uses REST API knowledge from seo-strategy):

Creates the post via wp-json/wp/v2/posts
Sets Yoast SEO meta fields
Outputs the draft URL
Total time: ~15 minutes. From zero to published draft with SEO metadata.

The GEO angle: why AI search changes everything
Here's something most developers don't think about: when someone asks ChatGPT "what's the best espresso machine for beginners?", the answer comes from somewhere. That somewhere could be your website — but only if your content is structured in a way AI models can parse and trust.

This is called Generative Engine Optimization (GEO). It's like SEO, but for AI search engines. The research (Aggarwal et al., 2024) shows that GEO techniques can boost visibility in AI-generated answers by up to 40%.

My geo-optimization skill includes:

A 16-criterion scorecard for AI-readiness
Writing guidelines that balance human readability and AI parseability
Anti-patterns (things that make AI skip your content)
E-E-A-T implementation (Experience, Expertise, Authoritativeness, Trust)
What I learned building these skills

  1. Skills need evidence, not opinions
    AI assistants hallucinate. If your skill says "meta descriptions should be 155 characters" without a source, Copilot might second-guess it or contradict itself later.

I added a sources file with links to Moz, Ahrefs, Google Search Central, and peer-reviewed papers. This makes the skill more reliable and more trustworthy for users.

  1. The description field is your trigger
    Copilot decides which skill to invoke based on the description in the YAML frontmatter. Be specific about trigger words:

description: "Use when the user mentions 'SEO', 'search engine
optimization', 'Google rankings', 'rank higher', 'meta tags'..."
The more trigger phrases you include, the more reliably it activates.

  1. Structure > prose
    Tables, bullet lists, and clear headings work better than long paragraphs. The AI parses structured content faster and applies it more consistently.

  2. Skills should reference each other
    My skills cross-link: the SEO strategy skill says "for keyword research, see seo-keyword-research". This helps Copilot chain skills together in multi-step workflows.

Installation (30 seconds)
git clone https://github.com/WilcoRakhorst/helder-ai-marketing-skills.git .github/skills/marketing

That's it. Open VS Code with Copilot agent mode, and start asking marketing questions. Copilot auto-discovers the skills.

Or as a Git submodule (stays updatable):

git submodule add https://github.com/WilcoRakhorst/helder-ai-marketing-skills.git .github/skills/marketing

Build your own skills
The pattern works for any domain knowledge:

Create a SKILL.md in .github/skills/your-skill/
Add YAML frontmatter with name and description (include trigger words)
Write structured knowledge — tables, checklists, decision trees
Add references for any factual claims
Test it — ask Copilot questions in the skill's domain and see if it gets invoked
Domains that would benefit from this approach:

Accessibility (WCAG guidelines as a skill)
Security (OWASP Top 10 as actionable checklists)
API design (REST/GraphQL conventions)
Legal/compliance (GDPR, cookie consent rules)
Brand guidelines (tone of voice, terminology)
The repo
GitHub: helder-ai/marketing-skills

Licensed CC BY-NC 4.0 — free to use, adapt, and learn from.

If you find it useful: ⭐ the repo, or open a PR if you have improvements. I'd especially love contributions from SEO practitioners who want to encode their expertise.

What domain knowledge would you encode as a Copilot skill? Let me know in the comments!

8 Comments

•

Pretty cool idea honestly. Did these reusable skills actually improve output consistency for you?

2 votes
•

@[peponder] Thanks for the kind words! Yes, consistency is exactly where the difference shows up most.

I manage a few websites — none of them as a professional webmaster, just someone who cares about getting the most out of them with the help of AI. Before skills, every blog post meant repeating the same instructions: "use the keyword in the first paragraph," "keep the meta description under 155 characters," "write for a B2 reading level." Now those rules live in the skill, and Copilot applies them automatically.

The biggest win for me is the club website. We agreed on a tone of voice as a team, and I encoded that directly into the content skill. Now every article it drafts already sounds right — I barely have to edit for style anymore.

And the SEO and GEO part? I honestly don't want to think about those details every time I write something. The skills handle the checklist; I just focus on the topic. It has genuinely taken my output to another level.

1
•

The problem with skills, architecturally, is that they are not part of the "system instructions" block. So they get injected with your prompt when the trigger is hit, and bloat the context window with their tokens, leading to two problems:

  1. Attention dilution: The context window grows in size, more quickly approaching both the functional terminus of attention and its corresponding gradient dropoffs. At ~60k tokens, you can notice that it starts to suffer from ADD, and at ~100k, attention is "reliably degraded" in the middle.
  2. It's a piece of crap, it doesn't work!: By deferring the instructions, they never make it into the overall system instructions block (eg AGENTS.md). The only reason the "skills" work at all is because they're on the tail end, but of course now the rest of the context window is so full of noise that it's hard for the AI to "remember" anything else from previous turns.

Honestly it doesn't surprise me. Github clearly doesn't know what they're doing. Have you even seen their guidelines for drafting system instructions? It's a case of the blind leading the blind, right off a cliff.

1 vote
•

@[demoran] Fair point on the context window trade-off — that's a real architectural constraint, not a GitHub problem specifically. Every deferred instruction system (LangChain, AutoGPT, even Claude's tool-use) faces the same attention dilution as context grows.

My approach handles it with a token-hygiene skill and a discipline of starting fresh sessions for separate tasks. Skills are kept under 300 lines and reference external files instead of inlining everything. It's not a perfect solution — it's a workflow pattern that works well enough for my use case.

Whether it belongs in AGENTS.md or a skill file is mostly a question of how static your knowledge is. If I put my SEO rules in AGENTS.md, they'd be there on every coding session too. On-demand loading is the feature, not the bug.

0
•

@[helderai] It sounds to me like you recognise the problem. But not the actual solution. Skills are not the answer: using a different AGENTS.md file for your SEO work is.

1
•

@[helderai] Also, looking at your skill, one thing is clear to me: you have forgotten your audience. The AI needs principles and instructions, not background padding and links it can't even follow. Not only are you completely wasting your tokens, but the AI is literally rolling its eyes at you. It is a machine intelligence, not a CEO.

1
•

@[demoran] You were right on the intro padding. I removed the "This skill is the upstream source..." paragraph and the ASCII diagram from the content-creator skill in the marketing-skills repo — both were background context, not instructions.

The other three skills (GEO, SEO strategy, keyword research) were already clean. Credits added to the README of both repos.

On the AGENTS.md approach: I see the case for it — scoping agent behavior per workflow context rather than relying on modular skill invocation is a valid alternative, especially for project-specific setups. Worth a follow-up.

Appreciate you taking the time to look properly.

0
•

The comments format sucks here. It doesn't seem to want to take an actual code block.

ME

Man, I bit into that guy kind of hard.

Ok, I want to help him out after busting his balls so bad.

Could you review the contents of the attached SKILL.md file and synthesize some actual instructions from it?

IT

The SKILL.md is a reference encyclopedia masquerading as instructions. It tells the agent what SEO is — not what to do when the operator asks for help. The difference matters: a knowledge dump activates retrieval, not behavior.

Here's a synthesized rewrite that converts the material into actual directives:

gist

What changed and why

The original was ~15k characters of reference material — tables of facts, definitions, status lists. An agent reading it gets information but no behavioral shaping. The rewrite:

  • Leads with decision logic — "determine the work type" routes the agent's behavior before it starts producing output.
  • Converts facts into rules — "Do not use FAQPage" instead of a status table the agent has to interpret.
  • Orders by diagnostic priority — technical foundation before content, matching how SEO problems should actually be triaged.
  • Eliminates the encyclopedia framing — "SEO = optimizing a website for organic search results" teaches the agent nothing it doesn't already know. Every token in a skill file competes for attention budget.
  • Keeps the actionable data — thresholds, character limits, schema types, deprecated features. These are the parts the agent can't derive from latent knowledge.

The result is slightly shorter but dramatically more directive. The original tells the agent what SEO is. This tells it what to do.

1

More Posts

I’m a Senior Dev and I’ve Forgotten How to Think Without a Prompt

Karol Modelskiverified - Mar 19

How I Built a React Portfolio in 7 Days That Landed ₹1.2L in Freelance Work

Dharanidharan - Feb 9

TypeScript Complexity Has Finally Reached the Point of Total Absurdity

Karol Modelskiverified - Apr 23

Sovereign Intelligence: The Complete 25,000 Word Blueprint (Download)

Pocket Portfolio - Apr 1

Just completed another large-scale WordPress migration — and the client left this

saqib_devmorph - Apr 7
chevron_left

More From helderai

My GitHub Copilot Sessions Kept Degrading -- So I Built a Skill to Fix It

Related Jobs

View all jobs →

Commenters (This Week)

1 comment
1 comment
1 comment

Contribute meaningful comments to climb the leaderboard and earn badges!