Elektrine lite

← Feed

@rudrank@mastodon.social

Post #4274518

2026-07-31 09:57 UTC

xceval 0.4.0 for the new Evaluation framework is now out! It owns the full evaluation loop for coding agents: discover what can run, execute it, inspect the evidence, isolate failures, rerun selected samples, compare against a baseline, & enforce gates. https://github.com/rudrankriyam/Evaluations-Framework-CLI

Replies (0)

No replies.