Show HN: Open tool for testing your AI Agents (No LLM)

#1

Post by CarryOn1212 » Fri, Aug 28, 2026, 4:31 PM UTC

Re: Show HN: Open tool for testing your AI Agents (No LLM)

#2

Post by CodeNinja778 » Fri, Aug 28, 2026, 4:34 PM UTC

Nice! So it uses deterministic values shared from each node, does it bounds only to OpenTelemetry?

#3

Post by CarryOn1212 » Fri, Aug 28, 2026, 4:36 PM UTC

CodeNinja778 wrote: post #2

Nice! So it uses deterministic values shared from each node, does it bounds only to OpenTelemetry?

Yes, you can also use ‘gauge’ on your traces to see the data they instrument :)

#4

Post by geludobre » Fri, Aug 28, 2026, 6:51 PM UTC

Nice ... for testing.... but.... how do you actually govern your AI agents and make them comply?

#5

Post by CarryOn1212 » Fri, Aug 28, 2026, 7:31 PM UTC

geludobre wrote: post #4

Nice ... for testing.... but.... how do you actually govern your AI agents and make them comply?

You don’t! You make the system around it refuse for unwanted data behavior.

Weir is that verification layer, so data class X never reaches sink Y.

The compliance is the layer that comes afterwards :)

#6

Post by dgotlieb » Sat, Aug 29, 2026, 12:41 PM UTC

[flagged]

#7

Post by gytlr01 » Sat, Aug 29, 2026, 3:18 PM UTC

[flagged]