Curation v2 step 5: deep assessment, utility, diversified shortlist

assess.rs replaces score.rs (DEEP_INSTRUCTIONS, representative sample with [BEGINNING]/[MIDDLE]/[END], facets, cached deep rows with --rescore bypass), rank.rs adds the utility blend over present signals with gate ramps and the leader-clustered shortlist (cap 2 → 3 → uncapped, protected top-N, exploration reserve), editor.rs replaces select.rs with the §13 rendering and utility-ordered fallbacks. ScoredArticle is gone; Candidate is the only flow type. deep_batch_size replaces score_batch_size. Started by Codex (cut off by its usage limit mid-verification) and finished by a Claude agent from docs/plans/curation-v2-briefs/step5.md; reviewed against plan §12–§13. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01A1rCLQeKBgnBo3oTgHuTMe
2026-09-02 16:04:43 +00:00
parent 10f4afd091
commit 05a74a0dcf
22 changed files with 2710 additions and 1298 deletions
@@ -0,0 +1,67 @@
+{
+  "articles": [
+    {
+      "id": 201,
+      "quality": 7.0,
+      "fit": 6.5,
+      "category": "Science & Space",
+      "rationale": "careful write-up of an amateur radio occultation measurement",
+      "paywalled_guess": false,
+      "facets": {
+        "format": "first_hand_account",
+        "depth": "standard",
+        "evidence": "data_or_experiment",
+        "commerciality": "none",
+        "topic_group": "science_space",
+        "technicality": "intermediate",
+        "locality": "us",
+        "specific_topics": ["radio occultation"]
+      }
+    },
+    {
+      "id": "202",
+      "quality": "6",
+      "fit": "5.5",
+      "category": "Niche Corner",
+      "rationale": "mailing-list argument about tape drives, oddly gripping",
+      "paywalled_guess": "false",
+      "facets": {
+        "format": "discussion_thread",
+        "depth": "standard",
+        "evidence": "anecdote",
+        "commerciality": "none",
+        "topic_group": "retro_computing",
+        "technicality": "intermediate",
+        "locality": "not_applicable",
+        "specific_topics": ["LTO tape", "backups", "archival", "vendors", "pricing"]
+      }
+    },
+    {
+      "id": 203,
+      "quality": 4,
+      "fit": 4
+    },
+    {
+      "id": 204,
+      "quality": 12.5,
+      "fit": -1,
+      "category": "Sports",
+      "rationale": "model ignored the rubric ceiling and invented a section here",
+      "paywalled_guess": false,
+      "facets": "not an object"
+    },
+    {
+      "id": 205,
+      "quality": 8.0,
+      "category": "Top Stories",
+      "rationale": "fit is missing, so this item is unusable"
+    },
+    {
+      "quality": 9.0,
+      "fit": 9.0,
+      "category": "Top Stories",
+      "rationale": "no id at all, unusable"
+    },
+    "a bare string where an object belongs"
+  ]
+}