AgentCapable scanner (agentcapable.dev)

What we do
  We request public web pages with agent-identifying User-Agent strings from
  two independent datacenter networks to measure how sites treat AI agents.
  Site owners and their tools use the results to fix silent agent blocking.

How we behave
  - Every scan reflects the site's current state; no long-lived result cache
    (only a 45-second double-submit window, bypassable with a re-scan).
  - We honor robots.txt for our own crawler identity (AgentCapable/0.1).
  - We pace requests (150ms gaps, ~40 requests per scan) and rate-limit per target.
  - We never submit forms, never log in, never go past a rate limit or CAPTCHA.
  - Scans are triggered by a person (or their agent) asking about a specific
    domain; we do not sweep the web.

Contact data
  Scans require name + email (phone optional). Stored with the scanned
  domain, used to attribute scans, email the report link, and send one
  monitoring-launch note.
  Never sold or shared. Deleted on request: abuse@agentcapable.dev

Opt out / abuse
  To exclude your site from scanning, or to report a problem:
  abuse@agentcapable.dev