regrun

5 ms·

1. 1.▲
   by regrun 8mo ago

   Great question. We haven’t published formal benchmarks yet, but in our demos we’re already blocking invalid or policy-violating calls before they hit downstream APIs (LLMs, payments, tools), which is where most marginal cost sits. Measuring
2. 2.▲
   Show HN: Reg.run - Decoupling AI "thinking" from API execution

   1 points by regrun 8mo ago|3 comments
3. 3.▲
   by regrun 8mo ago

   Thanks for the thoughtful questions! You've identified exactly the right challenges I am also facing, solution wise. > Where exactly does this sit? Between agent reasoning and tool execution. The agent/framework calls Reg.Run b
4. 4.▲
   Show HN: Reg.Run – Authorization layer for AI agents

   3 points by regrun 8mo ago|2 comments