fix: protect admin API endpoints; stop rejecting flood-peak measurements

Security: POST/PUT/DELETE /stations, POST /scrape/trigger and GET
/config now require an X-API-Key header matching ADMIN_API_KEY.
Secure by default - with no key configured those endpoints return 503
instead of being open. Comparison via secrets.compare_digest. Dashboard
and read endpoints stay public.

Data: the validator rejected any measurement whose discharge_percent
exceeded 200 - which silently deleted the Oct 2024 record-flood peaks
(the river genuinely ran at 201-226% of channel capacity). The cap is
now 500%, and an out-of-range percent nulls that auxiliary field
instead of discarding the whole row (water level and discharge are the
data that matter). Surfaced by the user's historical backfill log.

src/validators.py

@@ -22,7 +22,9 @@ class DataValidator:
     DISCHARGE_MIN = 0.0  # cms
     DISCHARGE_MAX = 10000.0  # cms
     DISCHARGE_PERCENT_MIN = 0.0
-    DISCHARGE_PERCENT_MAX = 200.0  # Allow some overflow
+    # % of channel capacity. Major floods genuinely exceed 200% (Oct 2024 peaked
+    # at 226%); values beyond 500% are treated as data errors.
+    DISCHARGE_PERCENT_MAX = 500.0
 
     @classmethod
     def validate_measurement(cls, measurement: Dict[str, Any]) -> bool:
@@ -59,7 +61,10 @@ class DataValidator:
                     logger.warning(f"Discharge out of range: {discharge}")
                     return False
 
-            # Validate discharge percent if present
+            # Validate discharge percent if present. This is an auxiliary field:
+            # an implausible value must not cost us the water level and discharge
+            # (rejecting rows here silently deleted the Oct 2024 flood peaks), so
+            # out-of-range percents are nulled and the measurement is kept.
             if measurement.get("discharge_percent") is not None:
                 discharge_percent = float(measurement["discharge_percent"])
                 if not (
@@ -68,9 +73,10 @@ class DataValidator:
                     <= cls.DISCHARGE_PERCENT_MAX
                 ):
                     logger.warning(
-                        f"Discharge percent out of range: {discharge_percent}"
+                        f"Discharge percent out of range ({discharge_percent}); "
+                        "keeping measurement with discharge_percent=None"
                     )
-                    return False
+                    measurement["discharge_percent"] = None
 
             # Validate station ID
             station_id = measurement["station_id"]

src/web_api.py

@@ -5,27 +5,38 @@ FastAPI web interface for water monitoring system
 
 import asyncio
 import os
+import secrets
 import time
 from contextlib import asynccontextmanager
 from datetime import datetime, timedelta
 from threading import Lock
-from typing import Any, Dict, List
+from typing import Any, Dict, List, Optional
 
 import requests
-from fastapi import BackgroundTasks, FastAPI, HTTPException, Query
+from fastapi import BackgroundTasks, Depends, FastAPI, Header, HTTPException, Query
 from fastapi.middleware.cors import CORSMiddleware
 from fastapi.responses import HTMLResponse
 from fastapi.staticfiles import StaticFiles
 
 from .config import Config
-from .health_check import (APIHealthCheck, DatabaseHealthCheck,
+from .health_check import (
-                           HealthCheckManager, MemoryHealthCheck)
+    APIHealthCheck,
+    DatabaseHealthCheck,
+    HealthCheckManager,
+    MemoryHealthCheck,
+)
 from .logging_config import get_logger, setup_logging
 from .metrics import get_metrics_collector, increment_counter, set_gauge
 from .postgres_history import PostgresHistory
-from .schemas import (HealthResponse, MeasurementResponse, MetricsResponse,
+from .schemas import (
-                      ScrapingStatusResponse, StationCreateModel,
+    HealthResponse,
-                      StationResponse, StationUpdateModel)
+    MeasurementResponse,
+    MetricsResponse,
+    ScrapingStatusResponse,
+    StationCreateModel,
+    StationResponse,
+    StationUpdateModel,
+)
 from .thaiwater import ThaiWaterClient
 from .water_scraper_v3 import EnhancedWaterMonitorScraper
 
@@ -40,6 +51,23 @@ FORECAST_CACHE: Dict[str, tuple] = {}
 FORECAST_CACHE_LOCK = Lock()
 FORECAST_TTL = 900  # 15 minutes
 
+# Admin API protection. Read/dashboard endpoints stay public; anything that
+# mutates state or leaks configuration requires the X-API-Key header matching
+# ADMIN_API_KEY. Secure by default: with no key configured, those endpoints
+# are disabled entirely rather than open.
+ADMIN_API_KEY = os.getenv("ADMIN_API_KEY")
+
+
+def require_admin_key(x_api_key: Optional[str] = Header(None, alias="X-API-Key")):
+    if not ADMIN_API_KEY:
+        raise HTTPException(
+            status_code=503,
+            detail="Admin API disabled: ADMIN_API_KEY is not configured on the server",
+        )
+    if not x_api_key or not secrets.compare_digest(x_api_key, ADMIN_API_KEY):
+        raise HTTPException(status_code=401, detail="Invalid or missing X-API-Key")
+
+
 # Dashboard HTML is loaded once at import from src/static/dashboard.html.
 _DASHBOARD_HTML_PATH = os.path.join(
     os.path.dirname(os.path.abspath(__file__)), "static", "dashboard.html"
@@ -269,7 +297,11 @@ async def get_stations():
     return stations
 
 
-@app.post("/stations", response_model=StationResponse)
+@app.post(
+    "/stations",
+    response_model=StationResponse,
+    dependencies=[Depends(require_admin_key)],
+)
 async def create_station(station: StationCreateModel):
     """Create a new monitoring station"""
     increment_counter("api_requests", labels={"endpoint": "create_station"})
@@ -319,7 +351,11 @@ async def create_station(station: StationCreateModel):
         raise HTTPException(status_code=500, detail=str(e))
 
 
-@app.put("/stations/{station_id}", response_model=StationResponse)
+@app.put(
+    "/stations/{station_id}",
+    response_model=StationResponse,
+    dependencies=[Depends(require_admin_key)],
+)
 async def update_station(station_id: int, updates: StationUpdateModel):
     """Update an existing monitoring station"""
     increment_counter("api_requests", labels={"endpoint": "update_station"})
@@ -374,7 +410,7 @@ async def update_station(station_id: int, updates: StationUpdateModel):
         raise HTTPException(status_code=500, detail=str(e))
 
 
-@app.delete("/stations/{station_id}")
+@app.delete("/stations/{station_id}", dependencies=[Depends(require_admin_key)])
 async def delete_station(station_id: int):
     """Delete a monitoring station"""
     increment_counter("api_requests", labels={"endpoint": "delete_station"})
@@ -594,7 +630,7 @@ async def get_station_measurements(
         raise HTTPException(status_code=500, detail=str(e))
 
 
-@app.post("/scrape/trigger")
+@app.post("/scrape/trigger", dependencies=[Depends(require_admin_key)])
 async def trigger_scraping(background_tasks: BackgroundTasks):
     """Trigger manual data scraping"""
     increment_counter("api_requests", labels={"endpoint": "scrape_trigger"})
@@ -654,7 +690,7 @@ async def get_scraping_status():
     )
 
 
-@app.get("/config")
+@app.get("/config", dependencies=[Depends(require_admin_key)])
 async def get_config():
     """Get current configuration (sensitive data masked)"""
     increment_counter("api_requests", labels={"endpoint": "config"})