RAG reranker + manifest bucket fix — quality improvements from eval

RAG pipeline now includes a cross-encoder rerank step between retrieval
and generation. The LLM re-sorts top-K results by relevance before
they become context. Falls back to original order if model output is
unparseable (~5% with 7B models). Also improved the generation prompt
to be domain-aware ("staffing database") and request specific citations.

Fixed 4 catalog manifests with bucket="data" (pre-federation leftover)
that poisoned the entire DataFusion query context on startup. The
"users", "lab_trials", "meta_runs", and "new_candidates" datasets
now correctly reference bucket="primary". This bug was surfaced by
the quality evaluation pipeline — wouldn't have been found by
structural tests alone.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

data/_catalog/manifests/3c2579b4-f3f3-4875-95fa-58d8b49ad94c.json

@@ -4,7 +4,7 @@
   "schema_fingerprint": "68f2c0d7a3ceb0aaa3c17c64900704519c72d213161bc9e5179c42ee53f6d0df",
   "objects": [
     {
-      "bucket": "data",
+      "bucket": "primary",
       "key": "datasets/meta_runs.parquet",
       "size_bytes": 729773,
       "created_at": "2026-03-28T01:38:57.380576453Z"