evals
Run evals on Jetty
Runbook diffs · 98 runbooks · 14 lineages

What changed between two runbooks

One lineage was edited by hand (Jon's claude-code + Sonnet sequence, v1 to v11). The rest were auto-evolved by the orchestrator: each round embeds the previous best SVG and names the weakest axis. Pick a lineage and any two versions.

Loading…

Hand-picked transitions

Click a card to load that pair.

Want this kind of eval for your own agents?

Every run on this site is a Jetty runbook: versioned instructions, a pinned sandbox, and a trajectory you can inspect. Point Jetty at your task and get the same receipts.