Full pipeline
Analyze → rank → rewrite → schema in one command. Output is copy-paste ready markdown and JSON-LD.
Rank and get cited in AI search engines
E-GEO analyzes, scores, and rewrites your content so ChatGPT, Perplexity, Gemini, and Claude can crawl, understand, and cite it. Open source, MIT licensed, research-backed.
What is E-GEO?
Section titled “What is E-GEO?”E-GEO — open-source Generative Engine Optimization (GEO) & Answer Engine Optimization (AEO) toolkit (Python CLI + Claude Code skills), based on published GEO research (arXiv:2511.20867).
It is built for developers, technical marketers, and site owners who want their content ranked and cited by ChatGPT, Perplexity, Gemini, and Claude. Unlike audit-only tools, E-GEO runs a full pipeline — analyze, rank-simulate, rewrite, and generate JSON-LD schema — in one command. It differentiates on a reproducible evaluation harness (verify prompt quality yourself, offline, no API key), a continuous geo-loop mode that watches your pages over time, a research-backed methodology (arXiv:2511.20867, building on Princeton’s KDD 2024 GEO study), and distribution as Claude Code skills. Install with pip install -e . or npx skills add https://github.com/mverab/eGEOagents.
Quickstart
Section titled “Quickstart”# Standalone Python CLIgit clone https://github.com/mverab/eGEOagents.git && cd eGEOagentspip install -e .egeo optimize examples/sample-input.md --out-dir ./geo-output
# Or as Claude Code skillsnpx skills add https://github.com/mverab/eGEOagentsNo API key required to try it — every command honors GEO_EVAL_MOCK=1 for deterministic offline runs.
Reproducible evaluation
A built-in harness measures whether the rewriter actually moves content up in an LLM-simulated ranking. Same check runs in CI.
Continuous geo-loop
Opt-in loop mode watches domains over time with deterministic collectors and a persistent workspace ($EGEO_HOME).
Research-backed
Based on the E-GEO research paper (arXiv:2511.20867), building on the original Princeton GEO study (KDD 2024).
How E-GEO compares
Section titled “How E-GEO compares”Last verified: 2026-08-05. Star counts as of 2026-08.
| E-GEO | geo-optimizer-skill (Auriti-Labs) | |
|---|---|---|
| GitHub stars | 147 | ~617 |
| Content rewriting | Yes — full pipeline | Audit-focused |
| Scoring | 10 research-derived GEO features | 0–100 across 47 methods |
| Reproducible evaluation harness | Yes (offline, deterministic, runs in CI) | No |
| Continuous loop mode | Yes (egeo loop, persistent workspace) |
No |
| Peer-reviewed paper by the authors | Yes (arXiv:2511.20867) | Builds on KDD 2024 GEO research |
| Astro integration | No | Yes |
| Claude Code skills distribution | Yes (skills.sh) | No |
| License | MIT | MIT |
geo-optimizer-skill has more stars and more scoring methods. E-GEO differentiates on rewriting (not just auditing), a reproducible evaluation harness, continuous loop mode, and a research-backed methodology. Full honest breakdown: E-GEO vs geo-optimizer-skill.
Get started
Section titled “Get started”Read the Getting Started guide, or go straight to the source on GitHub — MIT licensed, 147 stars and 42 forks as of 2026-08.