geoskills — Open-Source GEO Optimization Suite

Make your website discoverable, citable, and recommendable by AI search engines.

geoskills is an open-source suite of agent skills that brings Generative Engine Optimization (GEO) directly into your coding workflow. Instead of switching between audit tools, spreadsheets, and manual fixes, you run a single command and get actionable results — diagnosis, fixes, and progress tracking — all without leaving your terminal.

Repository: github.com/Cognitic-Labs/geoskills

The Audit-Fix-Monitor Workflow

Most GEO tools stop at "here's what's wrong." geoskills covers the full optimization cycle:

 1. AUDIT          2. FIX               3. MONITOR
 ┌──────────┐     ┌──────────────┐     ┌──────────────┐
 │geo-audit │ ──► │geo-fix-*     │ ──► │geo-monitor   │
 │          │     │  - llmstxt   │     │              │
 │Score: 48 │     │  - schema    │     │Score: 48→72  │
 │14 issues │     │  - content   │     │+24 points    │
 └──────────┘     └──────────────┘     └──────────────┘
       │                                      │
       └──────── geo-compare ─────────────────┘
              (benchmark vs competitors)

Step 1 — Diagnose: geo-audit scans your site across 4 dimensions and produces a scored report with prioritized issues.

Step 2 — Fix: Three targeted fix skills generate ready-to-use assets for the most common gaps:

  • geo-fix-llmstxt — Creates llms.txt and llms-full.txt so AI crawlers understand your site structure
  • geo-fix-schema — Generates JSON-LD markup (Organization, Article, FAQ, HowTo) tailored to your business type
  • geo-fix-content — Rewrites passages to remove hedge language, add data support, and improve self-containment

Step 3 — Measure: geo-monitor re-audits and compares against your baseline, showing exactly which scores improved and by how much.

Bonus: geo-compare benchmarks your site against 2-3 competitors side by side, revealing where they lead and where you have an advantage.

Quick Start

Install all skills with one command:

npx skills add Cognitic-Labs/geoskills

Then run your first audit:

/geo-audit https://your-site.com

For individual skills or alternative install methods, see the GitHub README.

Skill Reference

Skill Command What it does
geo-audit /geo-audit <url> Full 4-dimension GEO audit with composite score and 30-day roadmap
geo-fix-llmstxt /geo-fix-llmstxt <url> Generate llms.txt / llms-full.txt for AI discoverability
geo-fix-schema /geo-fix-schema <url> Analyze existing schema and generate missing JSON-LD markup
geo-fix-content /geo-fix-content <url> Rewrite content for AI citability with before/after comparison
geo-compare /geo-compare <url1> <url2> [url3] Side-by-side GEO score comparison across competing sites
geo-monitor /geo-monitor <url> Re-audit and track score changes against a previous baseline

When to Use What

Scenario Skill
"Why can't AI find my site?" geo-audit — start here
"AI crawlers can't read my site structure" geo-fix-llmstxt
"My pages have no structured data" geo-fix-schema
"My content is never cited by AI" geo-fix-content
"How do I compare to competitors in AI search?" geo-compare
"Did my fixes actually improve anything?" geo-monitor

Compatibility

geoskills works with any AgentSkills-compatible coding agent:

  • Claude Code
  • OpenCode
  • OpenClaw
  • Codex CLI
  • Cursor
  • GitHub Copilot

Diagnostic vs. Measurement

geoskills tells you what to fix. AIvsRank measures how visible you actually are — tracking real mentions across ChatGPT, Claude, Perplexity, Gemini, and Google AI Overviews over time.

Together, they give you the complete picture: diagnose with geoskills, measure with AIvsRank.

Previous Error Codes Next geo-audit