GEOInflow
Self-hosted web and AI crawler analytics for Generative Engine Optimization (GEO). Track visitors, AI search referrals, crawl logs, content coverage, and conversions. Jointly published by 猎看科技 · 北京日报社智源GEO. MIT licensed.
- ai-analytics
- geo
- mit
- postgresql
- self-hosted
- web-analytics
- Stars
- 352
- Forks
- 71
- + today
- +3
- Created
- 1d
Ranking data as of October 26, 2026 (UTC).
Star History
Today, hour by hour
Overview
Self-hosted web and AI crawler analytics for Generative Engine Optimization (GEO). Track visitors, AI search referrals, crawl logs, content coverage, and conversions. Jointly published by 猎看科技 · 北京日报社智源GEO. MIT licensed. It ranks #101 on GitTiger, gaining +3 stars on October 26, 2026 (UTC).
The project is written in TypeScript and has 71 forks. It was created 1d ago.
What it is
GEOInflow is a self-hosted analytics package for teams that want visibility into both ordinary site visitors and the AI crawlers that index content for generative search. Rather than renting a hosted dashboard, operators run it on infrastructure they control, a design the description links to Generative Engine Optimization work. It is described as jointly published by 猎看科技 and 北京日报社智源GEO; metadata records the MIT license and TypeScript.
Capabilities described
The description lists visitor tracking, referrals arriving from AI search, crawl logs, content coverage and conversion measurement. The README preview adds detail worth flagging: the core analytics and crawler-identification layer is said to work on your own servers with no model provider key, browser tracking is cookieless by default, and optional recording, event properties and business integrations are bounded by the scope configured for each deployment. AI analysis stays off unless an operator enables it, at which point they choose the model connection and which site material may be shared. A "GEO Weight" metric is marked Beta and said to draw on qualified AI traffic.
Getting started
Deployment instructions were not supplied, so database setup, environment configuration and the AI configuration guide the preview references must be read in the repository README. PostgreSQL appears among the topics, but this profile cannot confirm requirements.
Gaps
No benchmarks, hosting recipe or version history were included, and the Beta label on GEO Weight means its output deserves caution.
git clone https://github.com/yaojingang/GEOInflow.git
cd GEOInflow
# see README for setupFrequently Asked Questions
Does GEOInflow need an API key from a model provider?
Not for its core analytics and crawler identification, which the README says run on your own hardware without a model key; AI analysis is opt-in and each deployment picks its own model connection.
Is visitor tracking cookie-based?
The README preview describes browser tracking as cookieless by default, with optional recording and event properties governed by the scope set for each deployment.
Who publishes GEOInflow and under what license?
The description says it is jointly published by 猎看科技 and 北京日报社智源GEO, and the metadata records an MIT license.
Is the GEO Weight metric production ready?
The README preview labels GEO Weight as Beta and says it relies on qualified AI traffic, so its figures should be treated as preliminary.