feat(lineage): hover tooltip + click-to-inspect schema panel

- New route GET /admin/pipeline/lineage/schema/<model> — returns JSON
  with columns+types (from information_schema for serving models),
  row count, upstream and downstream model lists. Validates model
  against _DAG to prevent arbitrary table access.
- Precomputes _DOWNSTREAM map at import time from _DAG.
- Lineage template: replaces minimal edge-highlight JS with full UX —
  hover triggers schema prefetch + floating tooltip (layer badge, top 4
  columns, "+N more" note); click opens 320px slide-in panel showing
  row count, full schema table, upstream/downstream dep lists.
  Dep items in panel are clickable to navigate between models.
  Schema responses are cached client-side to avoid repeat fetches.
  Staging/foundation models show "schema in lakehouse.duckdb only".

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

web/src/padelnomics/admin/pipeline_routes.py

@@ -767,6 +767,12 @@ async def pipeline_trigger_extract():
 
 # ── Lineage tab ───────────────────────────────────────────────────────────────
 
+# Compute downstream map once at import time (DAG is static).
+_DOWNSTREAM: dict[str, list[str]] = {n: [] for n in _DAG}
+for _name, _deps in _DAG.items():
+    for _dep in _deps:
+        _DOWNSTREAM.setdefault(_dep, []).append(_name)
+
 
 @bp.route("/lineage")
 @role_required("admin")
@@ -780,6 +786,67 @@ async def pipeline_lineage():
     )
 
 
+@bp.route("/lineage/schema/<model>")
+@role_required("admin")
+async def pipeline_lineage_schema(model: str):
+    """JSON: schema details for a lineage node.
+
+    Returns columns + types from information_schema (serving models only —
+    staging/foundation live in lakehouse.duckdb which the web app cannot open).
+    Row count is included for serving models when the table exists.
+    """
+    from quart import jsonify
+
+    from ..analytics import fetch_analytics
+
+    if model not in _DAG:
+        return jsonify({"error": "unknown model"}), 404
+
+    layer = _classify_layer(model)
+    upstream = _DAG[model]
+    downstream = _DOWNSTREAM.get(model, [])
+
+    row_count = None
+    columns: list[dict] = []
+
+    if layer == "serving":
+        col_rows = await fetch_analytics(
+            """
+            SELECT column_name, data_type, is_nullable
+            FROM information_schema.columns
+            WHERE table_schema = 'serving' AND table_name = ?
+            ORDER BY ordinal_position
+            """,
+            [model],
+        )
+        columns = [
+            {
+                "name": r["column_name"],
+                "type": r["data_type"],
+                "nullable": r["is_nullable"] == "YES",
+            }
+            for r in col_rows
+        ]
+        if columns:
+            # model is validated against _DAG keys — safe to interpolate
+            count_rows = await fetch_analytics(
+                f"SELECT count(*) AS n FROM serving.{model}"
+            )
+            if count_rows:
+                row_count = count_rows[0]["n"]
+
+    return jsonify(
+        {
+            "model": model,
+            "layer": layer,
+            "upstream": upstream,
+            "downstream": downstream,
+            "row_count": row_count,
+            "columns": columns,
+        }
+    )
+
+
 # ── Catalog tab ───────────────────────────────────────────────────────────────