Observed arrival · 2026-10-09
AgentProbe puts AI agents through adaptive chaos tests
A testing tool that runs an AI agent, introduces disruptions, and reports how it responds.
○Open
⊠Login
$Paid
†Ads
✦Pretty
●Pro
◎Niche
◉Human
⚑Risk
ƒJS
- For
- Developers building AI agents
- Worth noticing
- Each step POSTs {task, history} to the agent endpoint and expects either a tool call or a final answer.
Field notes
The setup asks users to describe rules such as refund windows, escalation conditions, and actions that require approval. AgentProbe sends task and history to the supplied endpoint on each step and expects either a tool call or a final answer. The page says real tool schemas can improve test accuracy, while the adaptive injector’s API costs are billed to the user’s Anthropic key.
Observed signals
Read the marks
Editorial observations of this landing page, not a rating.
○OpenPublic substance visible
⊠LoginAccess appeared gated
✦PrettyNotable craft visible
◎NicheUnusually specific use
One card from the complete issue