Sprint 6: report review workflow, CI/K8s, and client tabs.

Add coach review APIs, pose calibration thresholds, Gitea CI, Kubernetes skeleton, H5 practice page, and mini program tab bar.

Co-authored-by: Cursor <cursoragent@cursor.com>

apps/api/app/ai/calibration.py

@@ -0,0 +1,13 @@
+"""Configurable thresholds for pose metric calibration."""
+
+CALIBRATION = {
+    "head_forward_scale": 180.0,
+    "shoulder_asymmetry_scale": 220.0,
+    "pelvic_tilt_scale": 4.5,
+    "normal_max": 35,
+    "low_max": 55,
+    "medium_max": 75,
+    "review_required_min": 75,
+}
+
+METRIC_NAMES = ("头前伸", "高低肩", "骨盆倾斜")

apps/api/app/ai/pose_metrics.py

@@ -15,7 +15,7 @@ RIGHT_HIP = 24
 Landmark = dict[str, float]
 FrameLandmarks = dict[int, Landmark]
 
-METRIC_NAMES = ("头前伸", "高低肩", "骨盆倾斜")
+from app.ai.calibration import CALIBRATION, METRIC_NAMES
 
 
 def _point(frame: FrameLandmarks, idx: int) -> Landmark | None:
@@ -43,7 +43,7 @@ def score_head_forward(frame: FrameLandmarks) -> float | None:
         return None
     mid_x = (left["x"] + right["x"]) / 2
     offset = abs(nose["x"] - mid_x) / width
-    return min(100.0, max(0.0, offset * 180))
+    return min(100.0, max(0.0, offset * CALIBRATION["head_forward_scale"]))
 
 
 def score_shoulder_asymmetry(frame: FrameLandmarks) -> float | None:
@@ -53,7 +53,7 @@ def score_shoulder_asymmetry(frame: FrameLandmarks) -> float | None:
     if not left or not right or not width:
         return None
     diff = abs(left["y"] - right["y"]) / width
-    return min(100.0, max(0.0, diff * 220))
+    return min(100.0, max(0.0, diff * CALIBRATION["shoulder_asymmetry_scale"]))
 
 
 def score_pelvic_tilt(frame: FrameLandmarks) -> float | None:
@@ -63,15 +63,15 @@ def score_pelvic_tilt(frame: FrameLandmarks) -> float | None:
         return None
     angle = abs(math.degrees(math.atan2(right["y"] - left["y"], right["x"] - left["x"])))
     tilt = min(angle, 180 - angle)
-    return min(100.0, max(0.0, tilt * 4.5))
+    return min(100.0, max(0.0, tilt * CALIBRATION["pelvic_tilt_scale"]))
 
 
 def value_to_level(value: float) -> str:
-    if value < 35:
+    if value < CALIBRATION["normal_max"]:
         return "normal"
-    if value < 55:
+    if value < CALIBRATION["low_max"]:
         return "low"
-    if value < 75:
+    if value < CALIBRATION["medium_max"]:
         return "medium"
     return "high"
 
@@ -111,12 +111,16 @@ def build_screening_report(frames: list[FrameLandmarks]) -> dict:
 
     worst = max((m["value"] for m in metrics), default=0)
     risk_level = value_to_level(worst)
+    if worst >= CALIBRATION["review_required_min"]:
+        risk_level = "review_required"
     if risk_level == "normal":
         summary = "体态指标整体正常，建议保持日常活动与姿势习惯"
     elif risk_level == "low":
         summary = "存在轻度体态偏差，建议开始基础纠正训练"
     elif risk_level == "medium":
         summary = "建议关注头前伸与高低肩，开始针对性训练"
+    elif risk_level == "review_required":
+        summary = "指标偏高，已提交机构教练复核"
     else:
         summary = "多项指标偏高，建议尽快安排专业评估与干预"