Add async pipeline with progress monitoring, resumability, and result transparency

Pipeline engine rewritten with combo-first loop: each combination is processed
through all requested passes before moving to the next, with incremental DB
saves after every step (crash-safe). Blocked combos now get result rows so they
appear in the results page with constraint violation reasons.

New pipeline_runs table tracks run lifecycle (pending/running/completed/failed/
cancelled). Web route launches pipeline in a background thread with its own DB
connection. HTMX polling partial shows live progress with per-pass breakdown.

Also: status guard prevents reviewed->scored downgrade, save_combination loads
existing status on dedup for correct resume, per-metric scores show domain
bounds + units + position bars, ensure_metric backfills units on existing rows.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

src/physcom/db/repository.py

@@ -3,7 +3,9 @@
 from __future__ import annotations
 
 import hashlib
+import json
 import sqlite3
+from datetime import datetime, timezone
 from typing import Sequence
 
 from physcom.models.entity import Dependency, Entity
@@ -170,6 +172,11 @@ class Repository:
             "INSERT OR IGNORE INTO metrics (name, unit, description) VALUES (?, ?, ?)",
             (name, unit, description),
         )
+        if unit:
+            self.conn.execute(
+                "UPDATE metrics SET unit = ? WHERE name = ? AND (unit IS NULL OR unit = '')",
+                (unit, name),
+            )
         row = self.conn.execute("SELECT id FROM metrics WHERE name = ?", (name,)).fetchone()
         self.conn.commit()
         return row["id"]
@@ -181,7 +188,7 @@ class Repository:
         )
         domain.id = cur.lastrowid
         for mb in domain.metric_bounds:
-            metric_id = self.ensure_metric(mb.metric_name)
+            metric_id = self.ensure_metric(mb.metric_name, unit=mb.unit)
             mb.metric_id = metric_id
             self.conn.execute(
                 """INSERT INTO domain_metric_weights
@@ -233,10 +240,13 @@ class Repository:
         combination.hash = self.compute_hash(entity_ids)
 
         existing = self.conn.execute(
-            "SELECT id FROM combinations WHERE hash = ?", (combination.hash,)
+            "SELECT id, status, block_reason FROM combinations WHERE hash = ?",
+            (combination.hash,),
         ).fetchone()
         if existing:
             combination.id = existing["id"]
+            combination.status = existing["status"]
+            combination.block_reason = existing["block_reason"]
             return combination
 
         cur = self.conn.execute(
@@ -255,6 +265,13 @@ class Repository:
     def update_combination_status(
         self, combo_id: int, status: str, block_reason: str | None = None
     ) -> None:
+        # Don't downgrade 'reviewed' to 'scored' — preserve human review state
+        if status == "scored":
+            row = self.conn.execute(
+                "SELECT status FROM combinations WHERE id = ?", (combo_id,)
+            ).fetchone()
+            if row and row["status"] == "reviewed":
+                return
         self.conn.execute(
             "UPDATE combinations SET status = ?, block_reason = ? WHERE id = ?",
             (status, block_reason, combo_id),
@@ -327,7 +344,7 @@ class Repository:
     def get_combination_scores(self, combo_id: int, domain_id: int) -> list[dict]:
         """Return per-metric scores for a combination in a domain."""
         rows = self.conn.execute(
-            """SELECT cs.*, m.name as metric_name
+            """SELECT cs.*, m.name as metric_name, m.unit as metric_unit
                FROM combination_scores cs
                JOIN metrics m ON cs.metric_id = m.id
                WHERE cs.combination_id = ? AND cs.domain_id = ?""",
@@ -335,12 +352,52 @@ class Repository:
         ).fetchall()
         return [dict(r) for r in rows]
 
-    def count_combinations_by_status(self) -> dict[str, int]:
-        rows = self.conn.execute(
-            "SELECT status, COUNT(*) as cnt FROM combinations GROUP BY status"
-        ).fetchall()
+    def count_combinations_by_status(self, domain_name: str | None = None) -> dict[str, int]:
+        """Count combos by status. If domain_name given, only combos with results in that domain."""
+        if domain_name:
+            rows = self.conn.execute(
+                """SELECT c.status, COUNT(*) as cnt
+                   FROM combination_results cr
+                   JOIN combinations c ON cr.combination_id = c.id
+                   JOIN domains d ON cr.domain_id = d.id
+                   WHERE d.name = ?
+                   GROUP BY c.status""",
+                (domain_name,),
+            ).fetchall()
+        else:
+            rows = self.conn.execute(
+                "SELECT status, COUNT(*) as cnt FROM combinations GROUP BY status"
+            ).fetchall()
         return {r["status"]: r["cnt"] for r in rows}
 
+    def get_pipeline_summary(self, domain_name: str) -> dict | None:
+        """Return a summary of results for a domain, or None if no results."""
+        row = self.conn.execute(
+            """SELECT COUNT(*) as total,
+                      AVG(cr.composite_score) as avg_score,
+                      MAX(cr.composite_score) as max_score,
+                      MIN(cr.composite_score) as min_score,
+                      MAX(cr.pass_reached) as last_pass
+               FROM combination_results cr
+               JOIN domains d ON cr.domain_id = d.id
+               WHERE d.name = ?""",
+            (domain_name,),
+        ).fetchone()
+        if not row or row["total"] == 0:
+            return None
+        # Also count blocked combos (they have no results but exist)
+        blocked = self.conn.execute(
+            "SELECT COUNT(*) as cnt FROM combinations WHERE status = 'blocked'"
+        ).fetchone()
+        return {
+            "total_results": row["total"],
+            "avg_score": row["avg_score"],
+            "max_score": row["max_score"],
+            "min_score": row["min_score"],
+            "last_pass": row["last_pass"],
+            "blocked": blocked["cnt"] if blocked else 0,
+        }
+
     def get_result(self, combo_id: int, domain_id: int) -> dict | None:
         """Return a single combination_result row."""
         row = self.conn.execute(
@@ -412,3 +469,88 @@ class Repository:
                 "pass_reached": r["pass_reached"],
             })
         return results
+
+    # ── Pipeline Runs ────────────────────────────────────────
+
+    def create_pipeline_run(self, domain_id: int, config: dict) -> int:
+        """Create a new pipeline_run record. Returns the run id."""
+        cur = self.conn.execute(
+            """INSERT INTO pipeline_runs (domain_id, status, config, created_at)
+               VALUES (?, 'pending', ?, ?)""",
+            (domain_id, json.dumps(config), datetime.now(timezone.utc).isoformat()),
+        )
+        self.conn.commit()
+        return cur.lastrowid
+
+    def update_pipeline_run(self, run_id: int, **fields) -> None:
+        """Update arbitrary fields on a pipeline_run."""
+        if not fields:
+            return
+        set_clause = ", ".join(f"{k} = ?" for k in fields)
+        values = list(fields.values())
+        values.append(run_id)
+        self.conn.execute(
+            f"UPDATE pipeline_runs SET {set_clause} WHERE id = ?", values
+        )
+        self.conn.commit()
+
+    def get_pipeline_run(self, run_id: int) -> dict | None:
+        row = self.conn.execute(
+            "SELECT * FROM pipeline_runs WHERE id = ?", (run_id,)
+        ).fetchone()
+        return dict(row) if row else None
+
+    def list_pipeline_runs(self, domain_id: int | None = None) -> list[dict]:
+        if domain_id is not None:
+            rows = self.conn.execute(
+                """SELECT pr.*, d.name as domain_name
+                   FROM pipeline_runs pr
+                   JOIN domains d ON pr.domain_id = d.id
+                   WHERE pr.domain_id = ?
+                   ORDER BY pr.created_at DESC""",
+                (domain_id,),
+            ).fetchall()
+        else:
+            rows = self.conn.execute(
+                """SELECT pr.*, d.name as domain_name
+                   FROM pipeline_runs pr
+                   JOIN domains d ON pr.domain_id = d.id
+                   ORDER BY pr.created_at DESC"""
+            ).fetchall()
+        return [dict(r) for r in rows]
+
+    def get_combo_pass_reached(self, combo_id: int, domain_id: int) -> int | None:
+        """Return the pass_reached for a combo in a domain, or None if no result."""
+        row = self.conn.execute(
+            """SELECT pass_reached FROM combination_results
+               WHERE combination_id = ? AND domain_id = ?""",
+            (combo_id, domain_id),
+        ).fetchone()
+        return row["pass_reached"] if row else None
+
+    def save_raw_estimates(
+        self, combo_id: int, domain_id: int, estimates: list[dict]
+    ) -> None:
+        """Save raw metric estimates (pass 2) with normalized_score=NULL.
+
+        Each dict: metric_id, raw_value, estimation_method, confidence.
+        """
+        for e in estimates:
+            self.conn.execute(
+                """INSERT OR REPLACE INTO combination_scores
+                   (combination_id, domain_id, metric_id, raw_value, normalized_score,
+                    estimation_method, confidence)
+                   VALUES (?, ?, ?, ?, NULL, ?, ?)""",
+                (combo_id, domain_id, e["metric_id"], e["raw_value"],
+                 e["estimation_method"], e["confidence"]),
+            )
+        self.conn.commit()
+
+    def get_existing_result(self, combo_id: int, domain_id: int) -> dict | None:
+        """Return the full combination_results row for resume logic."""
+        row = self.conn.execute(
+            """SELECT * FROM combination_results
+               WHERE combination_id = ? AND domain_id = ?""",
+            (combo_id, domain_id),
+        ).fetchone()
+        return dict(row) if row else None

src/physcom/db/schema.py

@@ -91,11 +91,29 @@ CREATE TABLE IF NOT EXISTS combination_results (
     UNIQUE(combination_id, domain_id)
 );
 
+CREATE TABLE IF NOT EXISTS pipeline_runs (
+    id              INTEGER PRIMARY KEY AUTOINCREMENT,
+    domain_id       INTEGER NOT NULL REFERENCES domains(id),
+    status          TEXT NOT NULL DEFAULT 'pending',
+    config          TEXT,
+    total_combos    INTEGER DEFAULT 0,
+    combos_pass1    INTEGER DEFAULT 0,
+    combos_pass2    INTEGER DEFAULT 0,
+    combos_pass3    INTEGER DEFAULT 0,
+    combos_pass4    INTEGER DEFAULT 0,
+    current_pass    INTEGER,
+    error_message   TEXT,
+    started_at      TIMESTAMP,
+    completed_at    TIMESTAMP,
+    created_at      TIMESTAMP DEFAULT CURRENT_TIMESTAMP
+);
+
 CREATE INDEX IF NOT EXISTS idx_deps_entity ON dependencies(entity_id);
 CREATE INDEX IF NOT EXISTS idx_deps_category_key ON dependencies(category, key);
 CREATE INDEX IF NOT EXISTS idx_combo_status ON combinations(status);
 CREATE INDEX IF NOT EXISTS idx_scores_combo_domain ON combination_scores(combination_id, domain_id);
 CREATE INDEX IF NOT EXISTS idx_results_domain_score ON combination_results(domain_id, composite_score DESC);
+CREATE INDEX IF NOT EXISTS idx_pipeline_runs_domain ON pipeline_runs(domain_id);
 """
 
 

src/physcom/engine/pipeline.py

@@ -1,15 +1,15 @@
-"""Multi-pass pipeline orchestrator."""
+"""Multi-pass pipeline orchestrator with incremental saves and resumability."""
 
 from __future__ import annotations
 
 from dataclasses import dataclass, field
+from datetime import datetime, timezone
 
 from physcom.db.repository import Repository
 from physcom.engine.combinator import generate_combinations
 from physcom.engine.constraint_resolver import ConstraintResolver, ConstraintResult
 from physcom.engine.scorer import Scorer
 from physcom.llm.base import LLMProvider
-from physcom.llm.prompts import PHYSICS_ESTIMATION_PROMPT, PLAUSIBILITY_REVIEW_PROMPT
 from physcom.models.combination import Combination, ScoredResult
 from physcom.models.domain import Domain
 
@@ -23,12 +23,17 @@ class PipelineResult:
     pass1_blocked: int = 0
     pass1_conditional: int = 0
     pass2_estimated: int = 0
+    pass3_scored: int = 0
     pass3_above_threshold: int = 0
     pass4_reviewed: int = 0
     pass5_human_reviewed: int = 0
     top_results: list[dict] = field(default_factory=list)
 
 
+class CancelledError(Exception):
+    """Raised when a pipeline run is cancelled."""
+
+
 def _describe_combination(combo: Combination) -> str:
     """Build a natural-language description of a combination."""
     parts = [f"{e.dimension}: {e.name}" for e in combo.entities]
@@ -53,158 +58,281 @@ class Pipeline:
         self.scorer = scorer
         self.llm = llm
 
+    def _check_cancelled(self, run_id: int | None) -> None:
+        """Raise CancelledError if the run has been cancelled."""
+        if run_id is None:
+            return
+        run = self.repo.get_pipeline_run(run_id)
+        if run and run["status"] == "cancelled":
+            raise CancelledError("Pipeline run cancelled")
+
+    def _update_run_counters(
+        self, run_id: int | None, result: PipelineResult, current_pass: int
+    ) -> None:
+        """Update pipeline_run progress counters in the DB."""
+        if run_id is None:
+            return
+        self.repo.update_pipeline_run(
+            run_id,
+            combos_pass1=result.pass1_valid
+            + result.pass1_conditional
+            + result.pass1_blocked,
+            combos_pass2=result.pass2_estimated,
+            combos_pass3=result.pass3_scored,
+            combos_pass4=result.pass4_reviewed,
+            current_pass=current_pass,
+        )
+
     def run(
         self,
         domain: Domain,
         dimensions: list[str],
         score_threshold: float = 0.1,
         passes: list[int] | None = None,
+        run_id: int | None = None,
     ) -> PipelineResult:
         if passes is None:
             passes = [1, 2, 3, 4, 5]
 
         result = PipelineResult()
 
+        # Mark run as running (unless already cancelled)
+        if run_id is not None:
+            run_record = self.repo.get_pipeline_run(run_id)
+            if run_record and run_record["status"] == "cancelled":
+                result.top_results = self.repo.get_top_results(domain.name, limit=20)
+                return result
+            self.repo.update_pipeline_run(
+                run_id,
+                status="running",
+                started_at=datetime.now(timezone.utc).isoformat(),
+            )
+
         # Generate all combinations
         combos = generate_combinations(self.repo, dimensions)
         result.total_generated = len(combos)
 
-        # Save all combinations to DB
+        # Save all combinations to DB (also loads status for existing combos)
         for combo in combos:
             self.repo.save_combination(combo)
 
-        # ── Pass 1: Constraint Resolution ───────────────────────
-        valid_combos: list[Combination] = []
-        if 1 in passes:
-            valid_combos = self._pass1_constraints(combos, result)
-        else:
-            valid_combos = combos
+        if run_id is not None:
+            self.repo.update_pipeline_run(run_id, total_combos=len(combos))
 
-        # ── Pass 2: Physics Estimation ──────────────────────────
-        estimated: list[tuple[Combination, dict[str, float]]] = []
-        if 2 in passes:
-            estimated = self._pass2_estimation(valid_combos, domain, result)
-        else:
-            # Skip estimation, use zeros
-            estimated = [(c, {}) for c in valid_combos]
+        # Prepare metric lookup
+        metric_names = [mb.metric_name for mb in domain.metric_bounds]
+        bounds_by_name = {mb.metric_name: mb for mb in domain.metric_bounds}
 
-        # ── Pass 3: Scoring & Ranking ───────────────────────────
-        scored: list[tuple[Combination, ScoredResult]] = []
-        if 3 in passes:
-            scored = self._pass3_scoring(estimated, domain, score_threshold, result)
+        # ── Combo-first loop ─────────────────────────────────────
+        try:
+            for combo in combos:
+                self._check_cancelled(run_id)
 
-        # ── Pass 4: LLM Review ──────────────────────────────────
-        if 4 in passes and self.llm:
-            self._pass4_llm_review(scored, domain, result)
+                # Check existing progress for this combo in this domain
+                existing_pass = self.repo.get_combo_pass_reached(
+                    combo.id, domain.id
+                ) or 0
 
-        # ── Save results after scoring ─────────────────────────
-        if 3 in passes:
-            max_pass = max(p for p in passes if p <= 5)
-            for combo, sr in scored:
-                self.repo.save_result(
-                    combo.id, domain.id, sr.composite_score,
-                    pass_reached=max_pass,
-                    novelty_flag=sr.novelty_flag,
-                    llm_review=sr.llm_review,
+                # Load existing result to preserve human review data
+                existing_result = self.repo.get_existing_result(
+                    combo.id, domain.id
                 )
-                self.repo.update_combination_status(combo.id, "scored")
 
-        # Collect top results
+                # ── Pass 1: Constraint Resolution ────────────────
+                if 1 in passes and existing_pass < 1:
+                    cr: ConstraintResult = self.resolver.resolve(combo)
+                    if cr.status == "blocked":
+                        combo.status = "blocked"
+                        combo.block_reason = "; ".join(cr.violations)
+                        self.repo.update_combination_status(
+                            combo.id, "blocked", combo.block_reason
+                        )
+                        # Save a result row so blocked combos appear in results
+                        self.repo.save_result(
+                            combo.id,
+                            domain.id,
+                            composite_score=0.0,
+                            pass_reached=1,
+                        )
+                        result.pass1_blocked += 1
+                        self._update_run_counters(run_id, result, current_pass=1)
+                        continue  # blocked — skip remaining passes
+                    elif cr.status == "conditional":
+                        combo.status = "valid"
+                        self.repo.update_combination_status(combo.id, "valid")
+                        result.pass1_conditional += 1
+                    else:
+                        combo.status = "valid"
+                        self.repo.update_combination_status(combo.id, "valid")
+                        result.pass1_valid += 1
+
+                    self._update_run_counters(run_id, result, current_pass=1)
+                elif 1 in passes:
+                    # Already pass1'd — check if it was blocked
+                    if combo.status == "blocked":
+                        result.pass1_blocked += 1
+                        continue
+                    else:
+                        result.pass1_valid += 1
+                else:
+                    # Pass 1 not requested; check if blocked from a prior run
+                    if combo.status == "blocked":
+                        result.pass1_blocked += 1
+                        continue
+
+                # ── Pass 2: Physics Estimation ───────────────────
+                raw_metrics: dict[str, float] = {}
+                if 2 in passes and existing_pass < 2:
+                    description = _describe_combination(combo)
+                    if self.llm:
+                        raw_metrics = self.llm.estimate_physics(
+                            description, metric_names
+                        )
+                    else:
+                        raw_metrics = self._stub_estimate(combo, metric_names)
+
+                    # Save raw estimates immediately (crash-safe)
+                    estimate_dicts = []
+                    for mname, rval in raw_metrics.items():
+                        mb = bounds_by_name.get(mname)
+                        if mb and mb.metric_id:
+                            estimate_dicts.append({
+                                "metric_id": mb.metric_id,
+                                "raw_value": rval,
+                                "estimation_method": "llm" if self.llm else "stub",
+                                "confidence": 1.0,
+                            })
+                    if estimate_dicts:
+                        self.repo.save_raw_estimates(
+                            combo.id, domain.id, estimate_dicts
+                        )
+
+                    result.pass2_estimated += 1
+                    self._update_run_counters(run_id, result, current_pass=2)
+                elif 2 in passes:
+                    # Already estimated — reload raw values from DB
+                    existing_scores = self.repo.get_combination_scores(
+                        combo.id, domain.id
+                    )
+                    raw_metrics = {
+                        s["metric_name"]: s["raw_value"] for s in existing_scores
+                    }
+                    result.pass2_estimated += 1
+                else:
+                    # Pass 2 not requested, use empty metrics
+                    raw_metrics = {}
+
+                # ── Pass 3: Scoring & Ranking ────────────────────
+                if 3 in passes and existing_pass < 3:
+                    sr = self.scorer.score_combination(combo, raw_metrics)
+
+                    # Persist per-metric scores with normalized values
+                    score_dicts = []
+                    for s in sr.scores:
+                        mb = bounds_by_name.get(s.metric_name)
+                        if mb and mb.metric_id:
+                            score_dicts.append({
+                                "metric_id": mb.metric_id,
+                                "raw_value": s.raw_value,
+                                "normalized_score": s.normalized_score,
+                                "estimation_method": s.estimation_method,
+                                "confidence": s.confidence,
+                            })
+                    if score_dicts:
+                        self.repo.save_scores(combo.id, domain.id, score_dicts)
+
+                    # Preserve existing human data
+                    novelty_flag = (
+                        existing_result["novelty_flag"] if existing_result else None
+                    )
+                    human_notes = (
+                        existing_result["human_notes"] if existing_result else None
+                    )
+
+                    self.repo.save_result(
+                        combo.id,
+                        domain.id,
+                        sr.composite_score,
+                        pass_reached=3,
+                        novelty_flag=novelty_flag,
+                        human_notes=human_notes,
+                    )
+                    self.repo.update_combination_status(combo.id, "scored")
+
+                    result.pass3_scored += 1
+                    if sr.composite_score >= score_threshold:
+                        result.pass3_above_threshold += 1
+
+                    self._update_run_counters(run_id, result, current_pass=3)
+                elif 3 in passes and existing_pass >= 3:
+                    # Already scored — count it
+                    result.pass3_scored += 1
+                    if existing_result and existing_result["composite_score"] is not None:
+                        if existing_result["composite_score"] >= score_threshold:
+                            result.pass3_above_threshold += 1
+
+                # ── Pass 4: LLM Review ───────────────────────────
+                if 4 in passes and self.llm:
+                    cur_pass = self.repo.get_combo_pass_reached(
+                        combo.id, domain.id
+                    ) or 0
+                    if cur_pass < 4:
+                        cur_result = self.repo.get_existing_result(
+                            combo.id, domain.id
+                        )
+                        if (
+                            cur_result
+                            and cur_result["composite_score"] is not None
+                            and cur_result["composite_score"] >= score_threshold
+                        ):
+                            description = _describe_combination(combo)
+                            db_scores = self.repo.get_combination_scores(
+                                combo.id, domain.id
+                            )
+                            score_dict = {
+                                s["metric_name"]: s["normalized_score"]
+                                for s in db_scores
+                                if s["normalized_score"] is not None
+                            }
+                            review = self.llm.review_plausibility(
+                                description, score_dict
+                            )
+
+                            self.repo.save_result(
+                                combo.id,
+                                domain.id,
+                                cur_result["composite_score"],
+                                pass_reached=4,
+                                novelty_flag=cur_result.get("novelty_flag"),
+                                llm_review=review,
+                                human_notes=cur_result.get("human_notes"),
+                            )
+                            result.pass4_reviewed += 1
+                            self._update_run_counters(
+                                run_id, result, current_pass=4
+                            )
+
+        except CancelledError:
+            if run_id is not None:
+                self.repo.update_pipeline_run(
+                    run_id,
+                    status="cancelled",
+                    completed_at=datetime.now(timezone.utc).isoformat(),
+                )
+            result.top_results = self.repo.get_top_results(domain.name, limit=20)
+            return result
+
+        # Mark run as completed
+        if run_id is not None:
+            self.repo.update_pipeline_run(
+                run_id,
+                status="completed",
+                completed_at=datetime.now(timezone.utc).isoformat(),
+            )
+
         result.top_results = self.repo.get_top_results(domain.name, limit=20)
         return result
 
-    def _pass1_constraints(
-        self, combos: list[Combination], result: PipelineResult
-    ) -> list[Combination]:
-        valid = []
-        for combo in combos:
-            cr: ConstraintResult = self.resolver.resolve(combo)
-            if cr.status == "blocked":
-                combo.status = "blocked"
-                combo.block_reason = "; ".join(cr.violations)
-                self.repo.update_combination_status(
-                    combo.id, "blocked", combo.block_reason
-                )
-                result.pass1_blocked += 1
-            elif cr.status == "conditional":
-                combo.status = "valid"
-                self.repo.update_combination_status(combo.id, "valid")
-                valid.append(combo)
-                result.pass1_conditional += 1
-            else:
-                combo.status = "valid"
-                self.repo.update_combination_status(combo.id, "valid")
-                valid.append(combo)
-                result.pass1_valid += 1
-        return valid
-
-    def _pass2_estimation(
-        self,
-        combos: list[Combination],
-        domain: Domain,
-        result: PipelineResult,
-    ) -> list[tuple[Combination, dict[str, float]]]:
-        metric_names = [mb.metric_name for mb in domain.metric_bounds]
-        estimated = []
-
-        for combo in combos:
-            description = _describe_combination(combo)
-            if self.llm:
-                raw_metrics = self.llm.estimate_physics(description, metric_names)
-            else:
-                # Stub estimation: derive from dependencies where possible
-                raw_metrics = self._stub_estimate(combo, metric_names)
-            estimated.append((combo, raw_metrics))
-            result.pass2_estimated += 1
-
-        return estimated
-
-    def _pass3_scoring(
-        self,
-        estimated: list[tuple[Combination, dict[str, float]]],
-        domain: Domain,
-        threshold: float,
-        result: PipelineResult,
-    ) -> list[tuple[Combination, ScoredResult]]:
-        scored = []
-        for combo, raw_metrics in estimated:
-            sr = self.scorer.score_combination(combo, raw_metrics)
-            if sr.composite_score >= threshold:
-                scored.append((combo, sr))
-                result.pass3_above_threshold += 1
-                # Persist per-metric scores
-                score_dicts = []
-                bounds_by_name = {mb.metric_name: mb for mb in domain.metric_bounds}
-                for s in sr.scores:
-                    mb = bounds_by_name.get(s.metric_name)
-                    if mb and mb.metric_id:
-                        score_dicts.append({
-                            "metric_id": mb.metric_id,
-                            "raw_value": s.raw_value,
-                            "normalized_score": s.normalized_score,
-                            "estimation_method": s.estimation_method,
-                            "confidence": s.confidence,
-                        })
-                if score_dicts:
-                    self.repo.save_scores(combo.id, domain.id, score_dicts)
-
-        # Sort by composite score descending
-        scored.sort(key=lambda x: x[1].composite_score, reverse=True)
-        return scored
-
-    def _pass4_llm_review(
-        self,
-        scored: list[tuple[Combination, ScoredResult]],
-        domain: Domain,
-        result: PipelineResult,
-    ) -> None:
-        for combo, sr in scored:
-            description = _describe_combination(combo)
-            score_dict = {s.metric_name: s.normalized_score for s in sr.scores}
-            review = self.llm.review_plausibility(description, score_dict)
-            sr.llm_review = review
-            result.pass4_reviewed += 1
-
     def _stub_estimate(
         self, combo: Combination, metric_names: list[str]
     ) -> dict[str, float]:
@@ -223,24 +351,21 @@ class Pipeline:
 
         # Rough speed estimate: F=ma -> v proportional to power/mass
         if "speed" in raw and mass_kg > 0:
-            # Very rough: speed ~ power / (mass * drag_coeff)
             raw["speed"] = min(force_watts / mass_kg * 0.5, 300000)
 
         if "cost_efficiency" in raw:
-            # Lower force = cheaper per km (roughly)
             raw["cost_efficiency"] = max(0.01, 2.0 - force_watts / 100000)
 
         if "safety" in raw:
-            raw["safety"] = 0.5  # default mid-range
+            raw["safety"] = 0.5
 
         if "availability" in raw:
             raw["availability"] = 0.5
 
         if "range_fuel" in raw:
-            # More power = more range (very rough)
             raw["range_fuel"] = min(force_watts * 0.01, 1e10)
 
         if "range_degradation" in raw:
-            raw["range_degradation"] = 365  # 1 year default
+            raw["range_degradation"] = 365
 
         return raw

src/physcom/models/domain.py

@@ -13,6 +13,7 @@ class MetricBound:
     weight: float  # 0.0–1.0
     norm_min: float  # Below this → score 0
     norm_max: float  # Above this → score 1
+    unit: str = ""
     metric_id: int | None = None
 
 

src/physcom/seed/transport_example.py

@@ -243,11 +243,11 @@ URBAN_COMMUTING = Domain(
     name="urban_commuting",
     description="Daily travel within a city, 1-50km range",
     metric_bounds=[
-        MetricBound("speed", weight=0.25, norm_min=5, norm_max=120),
-        MetricBound("cost_efficiency", weight=0.25, norm_min=0.01, norm_max=2.0),
-        MetricBound("safety", weight=0.25, norm_min=0.0, norm_max=1.0),
-        MetricBound("availability", weight=0.15, norm_min=0.0, norm_max=1.0),
-        MetricBound("range_fuel", weight=0.10, norm_min=5, norm_max=500),
+        MetricBound("speed", weight=0.25, norm_min=5, norm_max=120, unit="km/h"),
+        MetricBound("cost_efficiency", weight=0.25, norm_min=0.01, norm_max=2.0, unit="$/km"),
+        MetricBound("safety", weight=0.25, norm_min=0.0, norm_max=1.0, unit="0-1"),
+        MetricBound("availability", weight=0.15, norm_min=0.0, norm_max=1.0, unit="0-1"),
+        MetricBound("range_fuel", weight=0.10, norm_min=5, norm_max=500, unit="km"),
     ],
 )
 
@@ -255,11 +255,11 @@ INTERPLANETARY = Domain(
     name="interplanetary_travel",
     description="Travel between planets within a solar system",
     metric_bounds=[
-        MetricBound("speed", weight=0.30, norm_min=1000, norm_max=300000),
-        MetricBound("range_fuel", weight=0.30, norm_min=1e6, norm_max=1e10),
-        MetricBound("safety", weight=0.20, norm_min=0.0, norm_max=1.0),
-        MetricBound("cost_efficiency", weight=0.10, norm_min=1e3, norm_max=1e9),
-        MetricBound("range_degradation", weight=0.10, norm_min=100, norm_max=36500),
+        MetricBound("speed", weight=0.30, norm_min=1000, norm_max=300000, unit="km/s"),
+        MetricBound("range_fuel", weight=0.30, norm_min=1e6, norm_max=1e10, unit="km"),
+        MetricBound("safety", weight=0.20, norm_min=0.0, norm_max=1.0, unit="0-1"),
+        MetricBound("cost_efficiency", weight=0.10, norm_min=1e3, norm_max=1e9, unit="$/km"),
+        MetricBound("range_degradation", weight=0.10, norm_min=100, norm_max=36500, unit="days"),
     ],
 )