feat: serum stability in pipeline scores and pilot panel priority

src/openamp_foundry/pipeline.py

@@ -15,6 +15,7 @@
 from openamp_foundry.scoring.ensemble import ensemble_score, known_failure_modes, selection_reasons
 from openamp_foundry.scoring.novelty import novelty_score
 from openamp_foundry.scoring.safety import safety_score
+from openamp_foundry.scoring.stability import serum_stability_score
 from openamp_foundry.scoring.synthesis import synthesis_feasibility_score
 from openamp_foundry.selection.diversity import greedy_diverse_select
 from openamp_foundry.selection.pareto import rank_candidates
@@ -53,13 +54,15 @@ def score_candidates(
         synth = synthesis_feasibility_score(features, valid_sequence=valid)
         nov, nearest = novelty_score(candidate.sequence, references)
         boman_act = boman_activity_score(candidate.sequence) if valid else 0.0
+        stability = serum_stability_score(features) if valid else 0.0
         raw_scores = {
             "activity": act,
             "safety": safe,
             "synthesis": synth,
             "novelty": nov,
             "boman_activity": boman_act,
             "disagreement": model_disagreement(act, boman_act),
+            "serum_stability": stability,
         }
         raw_scores["ensemble"] = ensemble_score(raw_scores, weights)
         item = ScoredCandidate(

src/openamp_foundry/reports/pilot_panel.py

@@ -16,6 +16,7 @@
     "safety",
     "synthesis",
     "novelty",
+    "serum_stability",
     "pilot_priority",
 ]
 
@@ -48,6 +49,7 @@ def _f(key: str, fallback: float = 0.0) -> float:
         "safety": round(_f("safety"), 4),
         "synthesis": round(_f("synthesis"), 4),
         "novelty": round(_f("novelty"), 4),
+        "serum_stability": round(_f("serum_stability"), 4),
         "pilot_priority": round(_f("pilot_priority"), 4),
     }
 
@@ -86,21 +88,22 @@ def write_pilot_markdown(panel: list[dict], path: str | Path, generated_at: str
         "",
         "## Selection method",
         "",
-        "Priority score = `ensemble − 0.3 × disagreement`  ",
+        "Priority score = `ensemble − 0.3 × disagreement + 0.05 × serum_stability`  ",
         "Rules: one representative per seed (highest priority), then remaining slots filled by priority rank.",
+        "Serum stability acts as a light tiebreaker — favours candidates with fewer interior K/R cleavage sites.",
         "",
         "## Candidates",
         "",
-        "| # | ID | Sequence | Len | Seed | Ensemble | Activity | Boman | Disagree | Safety | Synth |",
-        "|--:|---|---|--:|---|---:|---:|---:|---:|---:|---:|",
+        "| # | ID | Sequence | Len | Seed | Ensemble | Activity | Boman | Disagree | Safety | Synth | SerumStab |",
+        "|--:|---|---|--:|---|---:|---:|---:|---:|---:|---:|---:|",
     ]
 
     for c in panel:
         r = _row(c)
         lines.append(
             f"| {r['pilot_rank']} | {r['candidate_id']} | `{r['sequence']}` | {r['length']} "
             f"| {r['seed']} | {r['ensemble']} | {r['activity']} | {r['boman_activity']} "
-            f"| {r['disagreement']} | {r['safety']} | {r['synthesis']} |"
+            f"| {r['disagreement']} | {r['safety']} | {r['synthesis']} | {r['serum_stability']} |"
         )
 
     lines += [

src/openamp_foundry/selection/pilot.py

@@ -14,10 +14,15 @@ def _seed_from_source(source: str) -> str:
 
 
 def _pilot_priority(scores: dict) -> float:
-    """Higher is better: reward ensemble score, penalise scorer disagreement."""
+    """Higher is better: reward ensemble, reward serum stability, penalise disagreement.
+
+    Serum stability is weighted lightly (0.05) so it acts as a tiebreaker without
+    overriding the ensemble score.  A candidate with stability=1.0 gains at most +0.05.
+    """
     ensemble = scores.get("ensemble", 0.0)
     disagreement = scores.get("disagreement", 0.5)
-    return round(ensemble - 0.3 * disagreement, 6)
+    stability = scores.get("serum_stability", 0.5)
+    return round(ensemble - 0.3 * disagreement + 0.05 * stability, 6)
 
 
 def select_pilot_panel(

tests/test_pilot_panel.py

@@ -81,8 +81,14 @@ def test_lower_disagreement_higher_priority(self):
         assert p1 > p2
 
     def test_missing_disagreement_defaults_to_0_5(self):
+        # stability defaults to 0.5 when absent; disagreement defaults to 0.5
         p = _pilot_priority({"ensemble": 0.70})
-        assert p == pytest.approx(0.70 - 0.3 * 0.5, abs=1e-6)
+        assert p == pytest.approx(0.70 - 0.3 * 0.5 + 0.05 * 0.5, abs=1e-6)
+
+    def test_higher_stability_higher_priority(self):
+        p1 = _pilot_priority({"ensemble": 0.70, "disagreement": 0.10, "serum_stability": 1.0})
+        p2 = _pilot_priority({"ensemble": 0.70, "disagreement": 0.10, "serum_stability": 0.0})
+        assert p1 > p2
 
 
 class TestSelectPilotPanel:
@@ -212,7 +218,7 @@ def test_correct_columns(self, tmp_path):
             fields = reader.fieldnames
         expected = {"pilot_rank", "candidate_id", "sequence", "length", "seed",
                     "ensemble", "activity", "boman_activity", "disagreement",
-                    "safety", "synthesis", "novelty", "pilot_priority"}
+                    "safety", "synthesis", "novelty", "serum_stability", "pilot_priority"}
         assert set(fields) == expected
 
     def test_row_count_matches_panel(self, tmp_path):