Leios prototype adds CallTrace tools and caps logging after Loki outages

What happened

The Ouroboros Leios prototype’s 2026w29 release adds CallTrace instrumentation to Forge and Mempool. The change is intended to provide more detailed visibility into time and space costs.

The same release caps FetchLogic state logging. The supplied record says these logs could reach tens of megabytes and had been breaking the Loki observability stack through limit breaches.

Why it matters

The release pairs work on Leios throughput with tools for measuring the path being changed. Cost profiling and stable telemetry are needed to assess whether prototype changes are helping, rather than relying on an unmeasured claim.

This is operational plumbing, but it affects how auditable later prototype performance results can be. If the telemetry system fails under load, it is harder to evaluate the measurements it produces.

What to watch next

Watch the prototype’s own throughput and voting dashboards for measured results, alongside evidence that the observability stack remains stable after the FetchLogic logging cap. Those are the next receipts for assessing whether the tooling supports reliable evaluation.