@hyperdx/hdx-eval@0.3.0

Minor Changes

• c4a2233: feat(evals): support an independent inference model/provider for LLM-as-judge
  grading. The grader can now differ from the run model (e.g. run with Anthropic,
  grade with OpenAI) to reduce same-model bias. --judge-model accepts a
  provider:model spec (anthropic, openai; a bare model name defaults to
  anthropic), configurable per repo via eval.config.json grading.judgeModel.
  The judge now uses the Vercel AI SDK, so grader credentials come from
  AI_API_KEY/OPENAI_API_KEY/ANTHROPIC_API_KEY (plus AI_BASE_URL /
  AI_REQUEST_HEADERS), mirroring the API package. The provider-specific key
  takes precedence over AI_API_KEY so a runner key and a differing grader key
  don't collide.

  The judge prompt now defines explicit 0-5 scoring anchors so scores stay
  calibrated across judge models, and the judge is more robust to intermittent
  structured-output malformations (recovers nested/stringified scores payloads
  and retries once on schema failure). Reasoning judge models (OpenAI gpt-5.x /
  o-series) get a larger output-token budget so hidden reasoning tokens don't
  truncate the JSON.

  fix(evals): re-grading a batch with a different judge model now actually
  re-runs that judge instead of silently returning the previously cached judge's
  scores. The needsJudge check and per-run reuse guard were keyed on the mere
  presence of a cached judge, not its identity, so grade <batch> --judge-model openai:gpt-5.6-sol over an Opus-graded batch would skip the LLM call and hand
  back the stale Opus scores. Both now key on the judge spec: a cached grade from
  a different judge model is treated as stale. --rerun-judge still forces a
  refresh with the same judge. This is required for grader-bias comparisons
  (grading one batch with two judges).

Patch Changes

• 9b2f9ce: feat(evals): harden the metric-saturation scenario
• fe5dfbb: feat(evals): grade & report metric-tool adoption
• cd1c2b2: feat(evals): add metrics-saturation scenario