Capture typed execution provenance for every agent run to make findings auditable and traceable to evidence

arXiv:2606.04990·low signal

A June 2026 survey on agent trust (arXiv:2606.04990) defines 'execution provenance' as the complete typed representation of an agent run — evidence units plus execution units — and 'evidence tracing' as the projection onto which evidence supported or influenced each output. This gives a vocabulary for building agents whose conclusions can be traced back to specific retrieved evidence. Actionable: log each agent step as a typed evidence/execution unit so any output can be replayed back to the exact sources and tool calls that produced it.