Auriti-Labs/geo-optimizer-skill

📦 geo‑optimizer‑skill – AI‑visibility audit & fix tool

What it is – An open‑source Python package (MIT‑licensed) that scores a website 0‑100 on how easily generative AI answer engines (ChatGPT, Perplexity, Gemini, Claude, Google AI Overviews) can crawl, understand, and cite it. It then produces concrete, prioritized fixes (robots.txt, llms.txt, JSON‑LD schema, meta tags, etc.).

Why it matters – Traditional SEO focuses on Google rankings, but LLM‑driven answer engines only cite a handful of “source‑ready” sites. This toolkit lets developers, SEO teams, or CI pipelines measure and improve that Answer Engine Optimization (AEO) / Generative Engine Optimization (GEO) readiness.

---

Quick start (zero‑install)

uvx --from geo-optimizer-skill geo audit --url https://example.com

Runs in seconds, no API keys required.

Or install normally:

pip install -U geo-optimizer-skill
geo audit --url https://example.com          # score + fixes
geo fix   --url https://example.com          # preview/write fix files
geo audit --url https://example.com --threshold 70   # exit 1 if score < 70 (CI gate)

---

Core features (as listed in the README)

Category	What the tool does
Scoring	0‑100 AI‑visibility score across 9 categories (robots.txt, llms.txt, schema, meta, content, brand, signals, AI discovery, Google‑AI readiness).
Fix generation	geo fix creates ready‑to‑publish files: robots.txt rules for 27 AI bots, llms.txt, JSON‑LD snippets, meta tags, etc.
CLI richness	16 commands (audit, fix, diff, history, drift, monitor, citations, snapshots, track, llms, schema, coherence, authority, logs, access, perception).
CI/CD integration	GitHub Action (Auriti-Labs/geo-optimizer-skill@vX) that fails a build below a configurable score and outputs SARIF/JUnit reports.
MCP server	Optional [mcp] extra installs a set of 12 “tool” endpoints usable from Claude, Cursor, Windsurf, etc.
Python API	from geo_optimizer import audit; result = audit(url) returns score, band, citability.total_score, breakdowns, and a list of recommendations.
Astro integration	Generates llms.txt, /.well-known/ai.txt, and /ai/summary.json at build time for static sites.
Monitoring & history	geo history, geo drift, geo track let you record scores over time, detect regressions, and receive email drift alerts (via the hosted geoready.dev service).
Citation checking	geo citations queries real answer engines (OpenAI, Anthropic, Perplexity, Gemini, SerpBase) to see if your brand/domain is cited, with multi‑run confidence intervals.
Extensibility	Plugin entry‑point (geo_optimizer.checks) for custom checks; example plugin provided.

---

Installation options

Method	Command	When to use
Zero‑install	uvx --from geo-optimizer-skill geo …	Quick one‑off audit, no Python env needed
Pip	pip install -U geo-optimizer-skill	Regular use, CI pipelines
Extras	pip install "geo-optimizer-skill[mcp]"	Need MCP tools for LLM agents

---

Typical workflow

1. Audit – geo audit --url https://yoursite.com → score + prioritized list.
2. Fix preview – geo fix --url … shows what files will be created.
3. Apply – add --apply to write the files, re‑run audit to see the new estimate.
4. CI gate – add the GitHub Action or run geo audit --threshold 70 in your pipeline.
5. Monitor – enable free hosted monitoring or run geo track --url … --report for periodic HTML reports.

---

License & community

• License: MIT – free for commercial or personal use.
• Package stats: ~2 000 + monthly downloads, CI badge shows active testing.
• Stars:  ⭐  on GitHub (as of the README).
• Documentation: /docs/ contains command guides, JSON contract, scoring rubric, and provider setup.

---

Bottom line

geo-optimizer-skill is a fully‑featured, open‑source engine that lets anyone quantify and improve how generative AI models discover and cite a website. It bridges SEO and LLM‑centric “search engineering,” offering both a CLI for ad‑hoc checks and integrations for CI, MCP agents, and static‑site generators.