production: harden catalog tenants analytics and widget install

ingest product variants (size/colour: per-variant price + stock) + bot-probe audit mode

coverage audit + scheduled daily re-ingest

catalog: sidebar-collection listing + reverse 'which category is X in'

eval panel: fix custom-tests 500, judge-key UX, product-question generation

fix: 'I specialize in [] for store' + eval relevance metric + mashup topics

fix: eval grader accepts variant-price answers containing reference price

feat: eval question diversity + OOS probes + score renormalization

fix(evals): add admin-only debug trace + evidence-based judge

fix(eval): correct product scoring and retrieval ranking

fix(eval): product-catalog question templates + judge rate-limit fixes

fix(eval): exclude retail logistics terms from non-retail DB tenant filter

fix(eval): two correctness fixes

fix(eval): use SQLite-first for document loading to avoid lock contention

fix: preserve rich config.json during GitHub sync; fix eval OOS filter for retail DBs

eval: derive product questions from chunks

eval: tighten tenant scope matching

eval: fix chat retrieval scoring and tenant scope fallback

eval: harden live run output and grounding filter

eval: tighten tenant scoping and fallback diagnosis

eval: tighten tenant filter and judge guidance

eval: target judge calls at failing rows

eval: confirm trace-backed judge verdicts locally

eval: deepen judge trace certainty

eval: add workflow-trace evidence to llm judge

eval: filter stale eval sets and diagnose idk rows

eval: filter cross-tenant analytics seeds

eval: add prompt-aware judge diagnostics

eval: add separate Groq judge layer

evals: add context recall and likely-cause hints

evals: improve live scoring and admin diagnostics

evals: add ranked retrieval and grounded answer metrics

evals: harden kb question quality and grading consistency

evals: remove cwd assumptions and expand coverage

evals: tighten grading and harden standalone suite

eval: harden DB-grounded suite and rotation

ci: run pytest and add eval v1

security: harden tenant auth and add lite eval