GEO project index

GitHub ai-robots-txt/ai.robots.txt

A community-curated registry of AI crawler user agents, published as ready-to-deploy robots.txt and web-server block rules generated from one JSON source.

Dataset Production-oriented Python · MIT · 4,069 · pushed Aug 22, 2026

GitHub angeo-dev/module-robots-txt-aeo

A Magento 2 module that prepends a managed AI-crawler block to robots.txt at render time, defaulting to Allow and leaving existing store rules untouched.

Library / SDK Production-oriented PHP · MIT · 5 · pushed Jun 14, 2026

GitHub AnswerDotAI/llms-txt

The canonical spec for /llms.txt, plus a reference Python CLI that flattens the file into a single LLM-ready context bundle.

Protocol / spec Usable Jupyter Notebook · Apache-2.0 · 2,578 · pushed Aug 10, 2026

GitHub dotenvx/llmstxt

Single-purpose CLI that walks a site's sitemap.xml, reads each page's title and description, and assembles them into an llms.txt index.

CLI Usable JavaScript · BSD-3-Clause · 147 · pushed Mar 5, 2026

GitHub galaar-org/AgentECHO

A dockerised stack of Next.js SDK, Go collector, ClickHouse and dashboard that records AI bot requests per route and estimates their licensing value.

Application Experimental TypeScript · 4 · pushed Jan 15, 2026

GitHub JerryZhi/AI-Crawler-Detector

A single-file Python probe that requests one URL once per AI crawler user agent and reports the HTTP status each identity is handed back.

CLI Experimental Python · MIT · 5 · pushed Aug 20, 2025

GitHub maango-io/ai-policy.json

A CC0 draft spec for one JSON manifest at /.well-known/ai-policy.json declaring whether AI systems may train on, search, or answer from a site.

Protocol / spec Experimental · CC0-1.0 · 5 · pushed Apr 18, 2026

GitHub madarco/ragrabbit

Self-hosted Next.js app that crawls your site into pgvector, then serves a reorderable llms.txt alongside an embeddable AI search widget and an MCP endpoint.

Application Experimental TypeScript · MIT · 136 · pushed Jan 5, 2026

⚠ Checked 2026-08-08: the default branch pins Next.js 15.2.3, and the fix for the January 2026 React Server Components advisory (15.2.8) was merged to a `demo` branch only. The README's one-click deploy therefore ships the unpatched version. Upgrade Next before exposing an instance.

GitHub MADEVAL/WP-robots-generator

A dependency-free browser page that assembles WordPress-aware robots.txt files, with one switch that writes Disallow for ten named AI crawlers.

Application Usable JavaScript · GPL-3.0 · 31 · pushed Jun 30, 2026

Hugging Face megrisdal/llms-txt

A CC0 corpus of 174 real llms.txt files with full text, collected from the community directory at directory.llmstxt.cloud via a published Kaggle notebook and maintained by Kaggle's product lead.

Dataset Usable HF Dataset · CC0-1.0 · 13 ♥ · 76 downloads/30d · updated Aug 17, 2026

GitHub nuxt-content/nuxt-llms

NuxtLabs-maintained Nuxt module that prerenders /llms.txt and an optional /llms-full.txt from configured sections plus content sources.

Library / SDK Production-oriented TypeScript · MIT · 167 · pushed Aug 21, 2026

GitHub pawamoy/mkdocs-llmstxt

MkDocs plugin that emits /llms.txt from configured section globs and converts each rendered page back to Markdown so generated API docs are included.

Library / SDK Production-oriented Python · ISC · 130 · pushed Aug 23, 2026

GitHub rubenmarcus/aeo.js

Build-time plugin set for a dozen JS frameworks that emits llms.txt, per-page Markdown, robots.txt, sitemap and JSON-LD, plus a CLI that scores live sites.

Library / SDK Experimental TypeScript · 129 · pushed Aug 11, 2026

⚠ No licence detected

GitHub stani-chirk/logwick

A self-hosted CLI and localhost dashboard that classifies edge JSONL access logs into humans, search crawlers, AI training fetches and AI user-fetch hits.

Application Usable JavaScript · 4 · pushed Jun 10, 2026

GitHub FabrizioCafolla/openai-crawlers-ip-ranges

A weekly cron-refreshed mirror of OpenAI's published GPTBot, ChatGPT-User and OAI-SearchBot fetch ranges, as bare addresses and CIDR.

Dataset Usable Shell · MIT · 31 · pushed Aug 17, 2026

GitHub firecrawl/llmstxt-generator

Firecrawl's now-deprecated web tool and API that crawled a site and compiled llms.txt and llms-full.txt files, retired in mid-2025 in favor of its main API.

Application Historical TypeScript · 537 · pushed Jun 17, 2025

⚠ No licence detected

GitHub langchain-ai/mcpdoc

LangChain's MCP server that lets IDEs and coding agents load user-chosen llms.txt indexes and fetch the pages they reference through auditable tool calls.

Library / SDK Usable Python · MIT · 1,030 · pushed Aug 20, 2026

GitHub JonasLong/Pyison

Self-hosted Python tarpit that traps rule-ignoring AI crawlers in an endless maze of procedurally generated pages, wasting their crawl budget on junk text.

Application Usable Python · MIT · 122 · pushed Oct 10, 2025

GitHub aaron-he-zhu/aaron-marketing-skills

A cross-agent marketing skill library whose SEO/GEO track moves from research and content implementation through technical checks, citation-readiness gates and performance monitoring.

Agent Skill Production-oriented Python · Apache-2.0 · 2,622 · pushed Aug 23, 2026

GitHub AEO-God-Mode/aeo-god-mode

A WordPress plugin that turns AI-search readiness into admin settings: crawler permissions, schema output, llms.txt, and a log of AI bot visits.

Library / SDK Production-oriented PHP · GPL-2.0 · 19 · pushed Aug 21, 2026

GitHub AgriciDaniel/claude-seo

A Claude Code plugin whose 25 skills and 18 agents audit a site across technical SEO, schema and AI-answer citability, then emit a prioritised action plan.

Agent Skill Production-oriented Python · MIT · 14,824 · pushed Aug 20, 2026

GitHub AgriciDaniel/codex-seo

A Codex-native SEO skill suite with twenty-six specialist workflows, TOML agent profiles and deterministic runners for audits, schema, GEO/AEO and reporting.

Agent Skill Production-oriented Python · 606 · pushed Jul 20, 2026

GitHub aigclink/geolook

A self-hosted dashboard that samples fifteen answer engines for brand mentions, turns the gaps into tickets and assets, then re-crawls to verify each fix.

Application Usable Python · MIT · 578 · pushed Aug 10, 2026

GitHub angeo-dev/module-rich-data

A Magento 2 module that injects Product, Organization, FAQPage and BreadcrumbList JSON-LD, plus the return and shipping fields AI shopping answers now expect.

Library / SDK Usable PHP · MIT · 5 · pushed Jun 14, 2026

GitHub Auriti-Labs/geo-optimizer-skill

Grades a site's AI-answer readiness across eight scoring categories, emits the concrete fixes, and then asks live engines whether the brand is actually cited.

Agent Skill Production-oriented Python · MIT · 730 · pushed Aug 17, 2026

Hugging Face cx-cmu/AutoGEO_Mini

The official demo Space for AutoGEO Mini: paste a document and get the GEO rewrite back in the browser, backed by the released checkpoints.

Application Research code HF Space · 1 ♥ · updated Apr 3, 2026

⚠ No licence detected

Hugging Face cx-cmu/AutoGEO_mini_Qwen1.7B_Ecommerce

AutoGEO Mini checkpoint specialized on the commercial e-commerce domain, trained against a gemini-2.5-flash-lite-powered engine for product-content rewriting.

Model Research code HF Model · MIT · 0 ♥ · 22 downloads/30d · updated Apr 11, 2026

Hugging Face cx-cmu/AutoGEO_mini_Qwen1.7B_GEOBench

AutoGEO Mini checkpoint specialized on the GEO-Bench domain, trained against a gemini-2.5-flash-lite-powered engine; the lab-organization variant paired with the GEO-Bench dataset configs.

Model Research code HF Model · MIT · 0 ♥ · 30 downloads/30d · updated Apr 11, 2026

Hugging Face cx-cmu/AutoGEO_mini_Qwen1.7B_ResearchyGEO

AutoGEO Mini checkpoint specialized on the Researchy-GEO domain: rewrites documents toward the preferences of a gemini-2.5-flash-lite-powered engine while aiming to preserve meaning and factual content.

Model Research code HF Model · MIT · 0 ♥ · 35 downloads/30d · updated Apr 11, 2026

GitHub dodopayments/dualmark

Gives every page a Markdown twin at the same URL, picked by HTTP content negotiation, backed by a public RFC-2119 spec and a conformance scorer for CI.

Library / SDK Usable TypeScript · Apache-2.0 · 101 · pushed Aug 22, 2026

GitHub geo-team-red/geo-optimizer

A pluggable Go library that rewrites content through five registrable strategies and scores it before and after with rule-based GEO checks.

Library / SDK Experimental Go · MIT · 190 · pushed Mar 27, 2026

GitHub IJONIS/geo-lint

A deterministic content linter whose JSON output is shaped for an agent to read, patch and re-run until clean, covering GEO, SEO, readability and i18n checks.

CLI Usable TypeScript · MIT · 39 · pushed Mar 19, 2026

GitHub jdevalk/seo-graph

A typed schema.org JSON-LD graph builder with an Astro integration that also serves Markdown twins, an llms.txt and RFC 9727 endpoints for agent discovery.

Library / SDK Usable TypeScript · MIT · 46 · pushed Aug 17, 2026

GitHub mverab/eGEOagents

A GEO toolkit in two runtimes, Claude Code agents and a Python CLI, that scores a page, rewrites it, emits JSON-LD, and can re-run as a weekly loop.

CLI Usable Python · MIT · 163 · pushed Aug 12, 2026

GitHub onvoyage-ai/gtm-engineer-skills

A Claude Code skill suite that crawls a site through sixteen foundational AEO checks, then turns six intelligence dimensions into framework-specific fixes.

Agent Skill Usable HTML · MIT · 1,280 · pushed Jun 7, 2026

GitHub rafters-studio/astro-meta

An Astro integration emitting what crawlers and LLMs read at build time: JSON-LD, llms.txt, an AI-crawler robots matrix, sitemap, OG images, and a scored audit.

Library / SDK Usable TypeScript · MIT · 6 · pushed Aug 3, 2026

GitHub rodrigostockebrand/oreilly-aeo-book

Companion code for an O'Reilly book on answer engine optimization, holding the full listings the book abbreviates: log parsers, JSON-LD, robots and llms.txt.

Template / example Usable Python · MIT · 1 · pushed Jul 16, 2026

GitHub seranking/seo-skills

A production-oriented Claude Agent Skills collection that turns SE Ranking data into content briefs, technical audits, structured data and AI-search share-of-voice reports.

Agent Skill Production-oriented Python · MIT · 123 · pushed Jun 25, 2026

GitHub spronta/crawlie

A Rust crawler that checks a site against 57 technical, accessibility and GEO rules, scoring AI-answer readiness per page and serving the same engine over MCP.

CLI Usable Rust · 99 · pushed Jul 18, 2026

GitHub yaojingang/GEORank

Self-hosted Chinese-language GEO workbench that audits a site's AI readiness, then turns the findings into JSON-LD, llms.txt, titles and a 30/60/90-day plan.

Application Usable Python · Apache-2.0 · 419 · pushed Aug 19, 2026

Hugging Face yujiangw/AutoGEO_mini_Qwen1.7B

The base AutoGEO Mini checkpoint: a Qwen3-1.7B model fine-tuned with GRPO reinforcement learning to rewrite web documents toward generative-engine preferences, at a reported fraction of frontier-model rewriting cost.

Model Research code HF Model · MIT · 0 ♥ · 15 downloads/30d · updated Dec 13, 2025

GitHub yulia-glukhova/aeo-schema-skill

An agent skill file walking Claude through Schema.org JSON-LD aimed at answer engines, built around a cross-page entity graph rather than per-page snippets.

Agent Skill Experimental · MIT · 2 · pushed Mar 13, 2026

GitHub zubair-trabzada/geo-seo-claude

A GEO-first Claude Code skill that audits a site, scores passage citability, checks AI crawler access and produces schema, llms.txt and platform-specific recommendations.

Agent Skill Production-oriented Python · MIT · 9,430 · pushed Aug 22, 2026

GitHub JayHoltslander/Structured-Data-JSON-LD

A hand-tested collection of schema.org JSON-LD examples across some thirty types, widely copied during the Google rich-results era and frozen since 2019.

Template / example Historical · 279 · pushed Nov 11, 2022

⚠ No licence detected

GitHub mustafapiplodi/schema-generator

A browser-only form builder for schema.org JSON-LD across 14 types, validating as you type and exporting as JSON, an HTML snippet, or GTM instructions.

Application Experimental JavaScript · 0 · pushed Nov 11, 2025

⚠ No licence detected

GitHub schemaorg/schemaorg

The upstream repository where the schema.org vocabulary is proposed, reviewed and built into the published site, under W3C community-group process since 2011.

Protocol / spec Production-oriented HTML · Apache-2.0 · 6,216 · pushed Aug 20, 2026

GitHub ai-search-guru/getcito-worlds-first-open-source-aio-aeo-or-geo-tool

A self-hosted monorepo that schedules prompts against AI answer engines via a Postgres job worker, then reports brand, competitor and cited-source share.

Application Production-oriented TypeScript · MIT · 170 · pushed Aug 22, 2026

GitHub aryamantodkar/oneglanse

Self-hosted tracker that opens your own logged-in sessions across five AI products in a browser and scores the answers they render, not API output, with an LLM.

Application Experimental TypeScript · MIT · 153 · pushed May 10, 2026

⚠ Checked 2026-08-08: capture works by signing into your own ChatGPT, Gemini, Perplexity, Claude and Google accounts and screen-scraping the rendered answers through an anti-fingerprint browser, chosen because stock automation hits bot checks. That deliberately circumvents anti-automation controls, very likely breaches those providers' terms, and risks suspension of whichever account you connect. Scores are also produced by an LLM judge over a single stochastic run, with no repeat sampling or confidence intervals. Its README claim to be the only free open-source GEO tracker is not correct.

GitHub Canonry/canonry

Self-hosted AEO platform that measures citations through provider APIs, crawler and referral traffic through server logs, then ships the fixes back to the site.

Application Production-oriented TypeScript · 117 · pushed Aug 20, 2026

GitHub danishashko/geo-aeo-tracker

A self-hosted Next.js dashboard that runs brand prompts across six AI answer surfaces and scores mention, citation and visibility drift over time.

Application Usable TypeScript · MIT · 244 · pushed Aug 12, 2026

GitHub elmohq/elmo

A self-hostable AI visibility tracker that runs prompt sets against answer engines and turns the responses into brand mention, citation, and competitor metrics.

Application Usable TypeScript · MIT · 258 · pushed Aug 22, 2026

GitHub onism1767-creator/potato

Local wizard and CLI that samples Claude's web-search answers on 24 frozen questions and reports brand mention and citation rates with confidence intervals.

CLI Experimental Python · MIT · 168 · pushed Jun 22, 2026

GitHub AIcling/agentic_geo

Self-evolving GEO rewriter: a quality-diversity strategy archive and a co-evolving critic pick a content-adaptive edit sequence from few engine calls.

Library / SDK Research code Python · Apache-2.0 · 39 · pushed Aug 5, 2026

Hugging Face cx-cmu/E-commerce

AutoGEO's commercial-domain dataset: product-oriented queries in the same six-config layout as its research domains, the smallest of the three (about 1.6k train, 400 test).

Dataset Research code HF Dataset · MIT · 2 ♥ · 222 downloads/30d · updated Dec 17, 2025

Hugging Face cx-cmu/GEO-Bench

AutoGEO's release of its research-domain GEO-Bench data, reorganized into six configs for the framework's pipeline: train/test, rule extraction, SFT cold start, inference, and GRPO input and eval (about 8k train, 1k test).

Dataset Research code HF Dataset · MIT · 1 ♥ · 117 downloads/30d · updated Dec 17, 2025

Hugging Face cx-cmu/Researchy-GEO

AutoGEO's research-domain evaluation data built on long, open-ended Researchy queries, in the same six-config layout as its sibling datasets and the largest of the three domains (about 10k train, 1k test).

Dataset Research code HF Dataset · MIT · 1 ♥ · 90 downloads/30d · updated Dec 17, 2025

GitHub cxcscmu/AutoGEO

Research code that mines a generative engine's content preferences into explicit rules, then rewrites documents with them via prompting or an RL-trained small model.

Library / SDK Research code Python · MIT · 202 · pushed Jun 14, 2026

GitHub EvoNexusX/2026LiuFeatGEO

Optimizes pages in an interpretable feature space (structure, content, language) via NSGA-II search, then has a model realize the winning config as prose.

Library / SDK Research code Python · 4 · pushed May 13, 2026

⚠ No licence detected

GitHub GEO-optim/GEO

Reference implementation and GEO-Bench evaluation harness from the paper that named Generative Engine Optimization, scoring how content edits shift source visibility.

Benchmark Research code Python · Apache-2.0 · 320 · pushed Oct 30, 2025

GitHub parameterlab/c-seo-bench

Companion code for a NeurIPS 2025 benchmark testing whether published conversational-SEO tactics lift citation rank across six domains and competing adopters.

Benchmark Research code Jupyter Notebook · MIT · 19 · pushed Sep 28, 2025

GitHub psbagga17/E-GEO

E-commerce GEO testbed: 13.7k long-form shopping queries over 86k products, scoring how far a query-blind rewrite lifts a product across five model judges.

Benchmark Research code Python · 45 · pushed Jul 3, 2026

⚠ No licence detected

Hugging Face psbagga17/E-GEO

The E-GEO testbed data: 13,747 long conversational shopping queries paired with ten Amazon listings each, plus the fixed 2,000-query test split, cached rankings from six model judges, and 3 GB of released run artifacts.

Dataset Research code HF Dataset · CC-BY-4.0 · 0 ♥ · 222 downloads/30d · updated Jul 3, 2026

Hugging Face qychen/CC-GSEO-Bench

The CC-GSEO-Bench data that the companion evaluation harness pulls from the Hub at runtime: 1,000+ source articles and 5,000+ query-article pairs for scoring GEO rewrite strategies on influence rather than citation alone.

Dataset Research code HF Dataset · Apache-2.0 · 0 ♥ · 35 downloads/30d · updated Dec 26, 2025

GitHub qychen2001/CC-GSEO-Bench

Benchmark harness that applies ten GEO rewrite strategies to an article, generates answers, then scores its influence with model judges on three metrics.

Benchmark Research code Python · 3 · pushed Dec 26, 2025

⚠ No licence detected

GitHub Wu-beining/MAGEO

Multi-agent GEO loop that profiles an engine's preferences, drafts several document revisions, then keeps only edits passing an attribution-fidelity gate.

Library / SDK Research code Python · 36 · pushed Jun 29, 2026

⚠ No licence detected

GitHub yaojingang/geo-citation-lab

Open citation-measurement workbench: cross-platform absorption experiments, a 214k-record Chinese AI-search citation dataset, and the pipelines behind both.

Dataset Research code HTML · 662 · pushed Aug 9, 2026

Hugging Face yonsei-dli/SAGEO-Arena

The SAGEO Arena release: 2,700 queries across nine domains sampled from established IR datasets, plus 261,874 Google Custom Search results; page content ships as URLs with crawler instructions rather than redistributed text.

Dataset Research code HF Dataset · MIT · 0 ♥ · 84 downloads/30d · updated May 28, 2026

GitHub zhizibianjie-omniedge/geo-cultural-encoding

Companion data sample and statistics scripts for a study of how training-data geography decides which brands a model will name at all.

Dataset Research code Python · MIT · 71 · pushed Jan 6, 2026

GitHub aounon/llm-rank-optimizer

Attack code from the strategic text sequence paper, optimizing an adversarial snippet on a product page so an LLM promotes that product to its top recommendation.

Library / SDK Research code Shell · MIT · 132 · pushed Aug 18, 2025

Hugging Face Bai-YT/RAGDOLL

The RAGDOLL dataset: 1,147 real consumer-product pages across ten categories for studying ranking manipulation in conversational search, shipped as a URL list plus downloaded pages for the paper's balanced subset.

Dataset Research code HF Dataset · CC-BY-4.0 · 4 ♥ · 1,809 downloads/30d · updated Jul 25, 2024

GitHub glad-lab/geobench

Benchmark that re-runs nine prior ranking-manipulation methods and ten white-hat strategies against one fixed LLM ranker, scoring effect and detectability.

Benchmark Research code Python · 3 · pushed Aug 3, 2026

⚠ No licence detected

GitHub glad-lab/MGEO

Code for an ACL 2026 workshop paper showing that coordinated image and text perturbations drive a chosen product to first place in a vision-language ranker.

Library / SDK Research code Python · 3 · pushed Jun 7, 2026

⚠ No licence detected

GitHub QianfengWen/SafeGEO

Benchmark that rewrites seller-controlled sources with seven GEO manipulation primitives, then measures how far recommendation agents drift from sound picks.

Benchmark Research code Python · Apache-2.0 · 5 · pushed Aug 22, 2026

GitHub spfrommer/cse-ranking-manipulation

Frozen artifact for the EMNLP 2024 paper showing that injected text can lift a low-ranked product in conversational search, including a live commercial engine.

Library / SDK Historical Python · 6 · pushed Oct 11, 2024

⚠ No licence detected

GitHub Tangyiming205069/controllable-seo

Official StealthRank code, showing that fluent adversarial text inside a product description can move an LLM ranker's ordering with no visible promotional cue.

Library / SDK Research code Python · 10 · pushed Jul 30, 2026

⚠ No licence detected

Hugging Face wieeii/SafeGEO

The SafeGEO benchmark data: 40,800 recommendation instances across six product verticals, pairing model-visible inputs with hidden labels for 22 GEO attack packages built from seven manipulation primitives.

Dataset Research code HF Dataset · CC-BY-4.0 · 0 ♥ · 203 downloads/30d · updated Jul 2, 2026

GitHub matanbt/GASLITE

Reproduction code for an ACM CCS 2025 attack that optimizes adversarial passages so dense-retrieval engines surface them for entire query concepts.

Library / SDK Research code Python · 14 · pushed Dec 6, 2025

⚠ No licence detected

GitHub yxf203/PoisonArena

Benchmark that runs six poisoning attacks against one another on a single RAG corpus, measuring which survive when several adversaries target the same query.

Benchmark Research code Python · 7 · pushed Jun 1, 2026

⚠ No licence detected

GitHub DavidHuji/Awesome-GEO

A single-file bibliography of GEO research organised by year, covering optimization methods and ranking-manipulation attacks plus a few open-source resources.

Reference collection Usable · 119 · pushed Aug 12, 2026

⚠ No licence detected

GitHub healsdata/ai-training-opt-out

A copy-paste reference of AI-training opt-out signals: robots.txt tokens, meta tags, HTTP headers, ai.txt, TDMRep, and crawler IP ranges, frozen in mid-2024.

Reference collection Historical HTML · Unlicense · 159 · pushed Jun 21, 2024

GitHub thedaviddias/llms-txt-hub

An actively maintained community directory that catalogues llms.txt implementations across the web, plus companion tooling for discovering and validating them.

Reference collection Usable TypeScript · 897 · pushed Aug 19, 2026

GitHub amplifying-ai/awesome-generative-engine-optimization

Large GEO reference directory indexing vendors, agencies, guides and papers rather than code, with a contribution queue stalled since April 2026.

Reference collection Usable · 488 · pushed Apr 14, 2026

⚠ No licence detected

GitHub luka2chat/awesome-geo

CC0 GEO handbook pairing a small link index with the maintainer's own checklists and templates, alongside large performance tables that cite almost nothing.

Reference collection Historical · CC0-1.0 · 135 · pushed Apr 9, 2026

⚠ Checked 2026-08-08: the document's own performance tables cite nothing. Figures such as a 340% visibility increase, a 92% mention rate, or a named company reaching a 78% AI citation rate carry no source, sample or date, and one claim asserts a measured effect from llms.txt that no major engine has confirmed consuming. Its arXiv citations do check out, but no performance number from this list should be quoted.