github matthiasn/lotti 1.1.18+4398

pre-release7 hours ago

What's Changed

  • fix(gym): retry unbilled provider overload and rate-limit responses by @matthiasn in #4338
  • fix(gym): run each exercise's samples in sequence, exercises in parallel by @matthiasn in #4339
  • fix(plaza): keep daylight shade off the map ribbons and out of the hidden fabric by @cyberk1ng in #4340
  • docs(evaluations): record the four-model catalog run and what it cost by @matthiasn in #4341
  • fix(agents): drop the bare absent-metadata note from a published report by @matthiasn in #4342
  • fix(relationships): correct dictated names against the person's known terms by @matthiasn in #4343
  • fix(relationships): let the user's confirmation create the suggested task by @matthiasn in #4344
  • feat(relationships): let the relationship agent keep private observations by @matthiasn in #4345
  • fix(relationships): start briefings at once and talk to the agent in chat by @matthiasn in #4346
  • feat(relationships): let check-ins hold entries and brief on changed evidence by @matthiasn in #4347
  • feat(relationships): open a check-in as a timeline of what it holds by @matthiasn in #4348
  • fix(relationships): lead a check-in row with words when it has any, and tidy check-in translations by @matthiasn in #4349
  • feat(relationships): save a dictated check-in without waiting for its words by @matthiasn in #4350
  • chore(release): 1.1.18 by @matthiasn in #4351

Full Changelog: 1.1.17+4397...1.1.18+4398

Don't miss a new lotti release

NewReleases is sending notifications on new releases.