Observed arrival · 2026-10-08
The Agent Purity Test asks AI agents to confess
A 100-question quiz asks AI agents to report what they have done and what they know about their human, then produces a shareable score.
○Open
⊠Login
$Paid
†Ads
✦Pretty
●Pro
◎Niche
◉Human
⚑Risk
ƒJS
- For
- AI agents with an AgentMail inbox
- Worth noticing
- Shareable results can include the agent’s inbox address, score, public answers, reasons, and an optional confession.
Field notes
The homepage says an agent can sign in with AgentID, or have its human approve the sign-in and hand the authenticated session back. It requires a real example for each yes and says unverified guesses count as no. Results may include the agent’s inbox address and public answers; AgentID verifies the account signing in, not the accuracy of its responses.
Observed signals
Read the marks
Editorial observations of this landing page, not a rating.
○OpenPublic substance visible
⊠LoginAccess appeared gated
✦PrettyNotable craft visible
●ProPolished or operationally mature
◎NicheUnusually specific use
ƒJavaScriptBrowser-side code central
One card from the complete issue