Skip to content

v0.0.9

Choose a tag to compare

View all tags
@github-actions github-actions released this 30 Sep 03:56
· 36 commits to trunk since this release
v0.0.9
This tag was signed with the committer’s verified signature.
rayandrew Ray Andrew
SSH Key Fingerprint: iGurnBY6QgoHsQWxP3NgvMEA4F3GjTcszIJnLk2jinw
Verified
Learn about vigilant mode.
6637ff0
This commit was signed with the committer’s verified signature.
rayandrew Ray Andrew
SSH Key Fingerprint: iGurnBY6QgoHsQWxP3NgvMEA4F3GjTcszIJnLk2jinw
Verified
Learn about vigilant mode.

Features

  • feat(distill): take 5 inbox items per run, configurable with [distill] batch (55405c6)
  • feat(hooks): inject only active lessons and pin the distill model (8e28370)
  • feat(observer): run each harness's own agent with a per-harness model chain (a0b67d4)
  • feat(rerank): let jev rate only the candidates it may see (8027688)
  • feat(observer): opt-in session observer that feeds durable notes to the inbox (37f848c)

Bug Fixes

  • fix(edit): tell a mistyped base hash apart from a changed lesson (9d0a958)
  • fix(hooks): keep the stop quiet on task signals and add failure_recall (8e44a8a)
  • fix(hooks): block the stop only for the user's own signals (b94965e)
  • fix(hooks): read signals and recall only from the text the user typed (6f15200)

Refactoring

  • refactor(config): share the config.toml loader and add [observer] (d763834)

Tests

  • test: agent A/B harness that runs tasks with rkb on and off (71b8f52)

Full Changelog: v0.0.8...v0.0.9

Assets 8
Loading