docs: refresh README + HF dataset card to current corpus (76,487 chunks, 11 sources)

tune: confirm retrieval history window = 2 turns (FE-managed multi-turn)

eval: faithfulness spot-check on health subset — no regression (5.00/5)

add scripts/test_api.py — CLI client to test /chat without a frontend

docs: update eval section with v7 results (citations 45/45, ~5s latency)

fix(safety): make _compute_needs_vet query required; pass query in eval harness

needs_vet: harden intent-gate with 20-case emergency stress set

needs_vet: add query intent-gate to cut emergency-banner over-trigger 8→1

feat(llm): migrate to google-genai SDK + quota resilience

docs(eval): record full judged eval — faithfulness 4.86, helpfulness 4.69

test(eval): score the shipped safety behavior (rerank gate + server banner)

fix(eval): make LLM judge robust to 2.5-flash thinking-token truncation

feat(llm): rotate across multiple Gemini API keys to multiply free-tier quota

fix(eval): use live Gemini models in judge fallback chain

feat(rag): add cross-encoder reranker, enforce citations, external-dataset eval harness

feat: harden safety flag, add API auth + rate limit, relevance gate, tests

fix: HF username is Monmoonluna (capital M), not monmoonluna

feat: HuggingFace Datasets upload/download scripts

feat: Windows Task Scheduler-based ingest watchdog

fix(finish_pipeline): strip Unicode em-dash that broke PowerShell parser

fix(finish_pipeline): drop Python sub-spawn for ChromaDB count

feat: ingest auto-restart wrapper + eval chain

feat: cache-hit short-circuit + auto-restart wrapper + omnibus pipeline

feat: 5 behavior-focused VN sources for phase 2 (3,383 new URLs)

test: eval script + 30-query bug-finding suite

test: standalone retrieval smoke-test script

fix(chunker): skip markdown table rows + separators as headings

feat: crawler for 5 Vietnamese cat data sources