Make AI model configurable per feature, and split news grading/formatting models

Every AI-calling feature (content correction, the @ai modifier, quiz grading,
SEO metadata generation, the AI Usage Analyzer, and DeepSearch) can now name
its own model via admin-editable configuration, defaulting to the gateway's
default model when left blank. The gateway a feature talks to stays fixed to
the internal endpoint either way, only the model name is a knob, so the best
model can be picked per task.

Also splits the news service's shared AI grading/formatting config into two
independent endpoint/model/key pairs: reformatting no longer requires
repointing (and thereby breaking) the free, scoring-only grading model.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FZ5x6KZTxZjbJqsEkxGbxG

tests/unit/services/jobs/deepsearch/worker.py

@@ -14,7 +14,7 @@ from devplacepy.services.deepsearch.store import VectorStore
 
 
 def _patch_pipeline(monkeypatch, pages):
-    async def fake_plan(query, api_key, emit=lambda frame: None):
+    async def fake_plan(query, api_key, emit=lambda frame: None, model=None):
         return [query, f"{query} overview"]
 
     async def fake_search(queries, emit=lambda frame: None):
@@ -29,7 +29,7 @@ def _patch_pipeline(monkeypatch, pages):
             outcome.pages.append(page)
         return outcome
 
-    async def fake_plan_followups(query, covered_titles, api_key, emit=lambda frame: None):
+    async def fake_plan_followups(query, covered_titles, api_key, emit=lambda frame: None, model=None):
         return []
 
     def fake_embed(texts, api_key):
@@ -38,7 +38,7 @@ def _patch_pipeline(monkeypatch, pages):
     async def fake_embed_async(texts, api_key, **kwargs):
         return local_embed(texts)
 
-    async def fake_orchestrate(question, crawled, api_key, emit, store=None, queries=None):
+    async def fake_orchestrate(question, crawled, api_key, emit, store=None, queries=None, model=None):
         from devplacepy.services.jobs.deepsearch.orchestrate import Orchestration
 
         return Orchestration(
@@ -166,7 +166,7 @@ def test_worker_run_performs_refinement_round_when_budget_remains(monkeypatch):
 
     followup_calls = []
 
-    async def fake_plan_followups_once(query, covered_titles, api_key, emit=lambda frame: None):
+    async def fake_plan_followups_once(query, covered_titles, api_key, emit=lambda frame: None, model=None):
         if followup_calls:
             return []
         followup_calls.append(covered_titles)

tests/unit/services/quiz/grading.py

@@ -134,7 +134,7 @@ def test_grade_free_text_uses_a_valid_verdict(monkeypatch):
 def test_grade_free_text_passes_the_answering_key_through(monkeypatch):
     seen = {}
 
-    def capture(api_key, system, text, timeout):
+    def capture(api_key, system, text, timeout, model=None):
         seen["api_key"] = api_key
         return json.dumps({"score": 1.0}), None