security: v1.7.7 hardening release

- Add WEBHOOK_CLIENT_SECRET validation for Graph webhooks
- Add Redis-backed rate limiting (fetch/ask/write/default tiers)
- Validate LLM_BASE_URL to prevent SSRF (HTTPS only, block private IPs)
- Enforce non-wildcard CORS when AUTH_ENABLED=true
- Add Content-Security-Policy headers
- Fix audit middleware to use verified JWT claims via contextvars
- Cap bulk_tags updates to 10,000 documents
- Return generic error messages to clients (no internal detail leakage)
- Strict AlertCondition Pydantic model for alert rules
- Security warning on MCP stdio server startup
- Remove MongoDB/Redis host ports from docker-compose
- Remove mongo_query from /ask API response

backend/routes/ask.py

@@ -397,8 +397,31 @@ def _format_events_for_llm(
     return "\n".join(lines)
 
 
+def _validate_llm_url(url: str):
+    """Prevent SSRF by rejecting internal/reserved addresses."""
+    from urllib.parse import urlparse
+
+    parsed = urlparse(url)
+    if parsed.scheme != "https":
+        raise RuntimeError("LLM_BASE_URL must use HTTPS")
+    hostname = (parsed.hostname or "").lower()
+    if not hostname:
+        raise RuntimeError("LLM_BASE_URL must have a valid hostname")
+    blocked = {"localhost", "127.0.0.1", "0.0.0.0", "::1", "169.254.169.254"}
+    if hostname in blocked:
+        raise RuntimeError(f"LLM_BASE_URL hostname '{hostname}' is not allowed")
+    # Block link-local and private IP ranges
+    import ipaddress
+
+    try:
+        ip = ipaddress.ip_address(hostname)
+        if ip.is_private or ip.is_loopback or ip.is_link_local or ip.is_reserved:
+            raise RuntimeError(f"LLM_BASE_URL IP '{hostname}' is not allowed")
+    except ValueError:
+        pass  # hostname is not an IP, which is fine
+
+
 def _build_chat_url(base_url: str, api_version: str) -> str:
-    """Construct the chat completions URL, handling Azure OpenAI endpoints."""
     base = base_url.rstrip("/")
     url = base if base.endswith("/chat/completions") else f"{base}/chat/completions"
     if api_version:
@@ -424,6 +447,9 @@ async def _call_llm(
         },
     ]
 
+    # SSRF guard: only allow known public HTTPS endpoints
+    _validate_llm_url(LLM_BASE_URL)
+
     url = _build_chat_url(LLM_BASE_URL, LLM_API_VERSION)
     headers = {
         "Content-Type": "application/json",
@@ -570,6 +596,8 @@ async def _explain_event(event: dict, related: list[dict]) -> str:
         },
     ]
 
+    _validate_llm_url(LLM_BASE_URL)
+
     url = _build_chat_url(LLM_BASE_URL, LLM_API_VERSION)
     headers = {"Content-Type": "application/json"}
     if "azure" in LLM_BASE_URL.lower() or "cognitiveservices" in LLM_BASE_URL.lower():
@@ -731,7 +759,7 @@ async def ask_question(body: AskRequest, user: dict = Depends(require_auth)):
         raw_events = list(cursor)
     except Exception as exc:
         logger.error("Failed to query events for ask", error=str(exc))
-        raise HTTPException(status_code=500, detail=f"Database query failed: {exc}") from exc
+        raise HTTPException(status_code=500, detail="Database query failed") from exc
 
     for e in raw_events:
         e["_id"] = str(e.get("_id", ""))
@@ -803,7 +831,6 @@ async def ask_question(body: AskRequest, user: dict = Depends(require_auth)):
                     "total_matched": total,
                     "services_queried": query_services,
                     "excluded_services": excluded_services,
-                    "mongo_query": json.dumps(query, default=str),
                 },
                 llm_used=False,
                 llm_error=None,
@@ -863,7 +890,6 @@ async def ask_question(body: AskRequest, user: dict = Depends(require_auth)):
             "total_matched": total,
             "services_queried": query_services,
             "excluded_services": excluded_services,
-            "mongo_query": json.dumps(query, default=str),
         },
         llm_used=llm_used,
         llm_error=llm_error,