github Agenta-AI/agenta v0.128.7

4 hours ago

What's Changed

  • fix(api): live evaluations fail every evaluator step and re-evaluate their own evaluator traces by @mmabrouk in #8101
  • perf(frontend): load the playground runner only on the pages that need it in /m by @mmabrouk in #8105
  • fix(frontend): report a saved app rename as a success when the read-back fails by @mmabrouk in #8096
  • [fix] Edit your own /m trace annotation and save only the set fields by @mmabrouk in #8102
  • fix(frontend): queue scan stops on a failed trace read, Retry sends only the rest, live filter drops annotation fields by @mmabrouk in #8098
  • fix(frontend): /m playground fixes before preview (phone header, JSON text type, commit version) by @mmabrouk in #8099
  • [feat] Run a scenario's invocation from the /m evaluation run page by @mmabrouk in #8103
  • fix(frontend): fix the endpoints snippet, prompts type, variants labels and the traces filter warning in /m by @mmabrouk in #8076
  • fix(api): let an app move from a folder back to the top level by @mmabrouk in #8077
  • feat(frontend): Playground share links, last selection and links in from other /m pages by @mmabrouk in #8075
  • fix(frontend): show a plain message when the /w queue add fails by @mmabrouk in #8112
  • [release] v0.128.7 by @github-actions[bot] in #8106

Full Changelog: v0.128.6...v0.128.7

Don't miss a new agenta release

NewReleases is sending notifications on new releases.