# Agent Web Lab

- Page: https://scanovich.ai/projects/agent-web-lab/
- Language: en
- Updated: 2026-08-10

- Status: research
- Code: open source · MIT
- Repository: https://github.com/FUYOH666/scanovich-agent-web-lab
- Research line: [The agent-readable web](https://scanovich.ai/research/agent-readable-web.md)

Open experiments on how AI agents discover, interpret, compare and act on what a business publishes on the web.

A measurement programme on a real commercial domain. Not a ranking playbook.

## What works today

- One pipeline measured stage by stage: fetch, retrieval, understanding, consideration, citation, action.
- Hypotheses registered before each run. Holdout sets stay sealed.
- A public repository with the method, the schemas and the scrubbed results.

## Why it is in the lab

People now send agents to read for them. This site is built from what the lab learned: one source of facts, served to people and to agents.

## Evidence

- **Method + results** — Preregistration, schemas and scrubbed results are public. Measured 2026-08-10. Source: https://github.com/FUYOH666/scanovich-agent-web-lab

## How to engage

Run the same measurement on your domain — what agents fetch, what they understand, what they cite.

[Contact](https://scanovich.ai/contact/) · private@scanovich.ai