github matthiasn/lotti 1.0.11+4333

latest release: 1.0.11+4336
pre-release4 hours ago

What's Changed

  • feat: nudge banner substrate by @cyberk1ng in #3944
  • fix(goals): unbreak main, pin rolling aggregates, measure the eval noise floor by @matthiasn in #3961
  • fix(evals): make the dialogue fixtures possible, and stop out-enforcing the runtime by @matthiasn in #3962
  • test(evals): score the clarifying question where it has to be by @matthiasn in #3964
  • test(evals): add a tier-2 goal eval that scores what the wake persisted by @matthiasn in #3965
  • fix(goals): tell the report every rule it broke, not just the first by @matthiasn in #3966
  • fix(evals): stop the outcome eval passing wakes that persisted no report by @matthiasn in #3967
  • feat(goals): give a goal a journal-side entity by @matthiasn in #3969
  • fix(goals): accept a reader-formatted aggregate as quoting the aggregate by @matthiasn in #3968
  • feat(goals): bind a goal agent to the goal it coaches by @matthiasn in #3970
  • feat(timeline): extract one shared timeline component and port events onto it by @matthiasn in #3972
  • feat(goals): write the journal-side goal whenever the spec changes by @matthiasn in #3973
  • feat(goals): capture check-ins and put them on the goal's timeline by @matthiasn in #3974
  • feat(goals): compact check-ins and feed the agent what the user said by @matthiasn in #3975
  • fix(goals): reconcile check-in summaries, and stop reading them twice by @matthiasn in #3976
  • docs(evals): record the open gaps and the provider probe by @matthiasn in #3977
  • fix: polish goals status and check-in UX by @matthiasn in #3978

Full Changelog: 1.0.11+4332...1.0.11+4333

Don't miss a new lotti release

NewReleases is sending notifications on new releases.