Non-blocking concerns raised while reviewing PR #15 (deps(plugin-eval): bump anthropic from 0.116.0 to 1.4.0 in /plugins/plugin-eval).
None of these blocked the merge. They are batched into one issue so a
review's findings stay one unit of attention rather than 1 separate
tracking issues; tick items off as they are addressed, and close this issue
when the list is done or the remaining items are judged not worth doing.
Non-blocking concerns raised while reviewing PR #15 (deps(plugin-eval): bump anthropic from 0.116.0 to 1.4.0 in /plugins/plugin-eval).
None of these blocked the merge. They are batched into one issue so a
review's findings stay one unit of attention rather than 1 separate
tracking issues; tick items off as they are addressed, and close this issue
when the list is done or the remaining items are judged not worth doing.
@dependabot rebaseby twistedmelonman; no reviewer explicitly flagged this),plugins/plugin-eval/uv.lock)This is a 0.x → 1.x major version bump of the
anthropicSDK, and the lockfile shows the transitive dependency set changing materially —distroandhttpxare dropped in favor ofhttpx2,httpcore2,truststore, andhttpx2-jsfetch, withidnamoving 3.15 → 3.19. No application source is modified in this PR, so any 1.x breaking API changes would only surface at runtime in paths the test suite does not cover. Per the repo's own auto-merge policy, majors are not auto-merged and get human review; suggested action is a quick pass over theanthropic1.0 migration notes and confirmation that plugin-eval's client-construction and streaming call sites still match the 1.x surface. The passingplugin-eval pytestand real-CLI smoke checks are supporting evidence but not proof of full API coverage.