Security audit

Frigate MQTT Camera Monitor and Notifier

Security checks for vulnerabilities and agentic risk

Overview

This skill does what it claims: it watches Frigate events and sends user-configured WhatsApp and email alerts, including sensitive camera snapshots.

Install only if you intend to send Frigate detection details and camera images to the configured WhatsApp number and email address. Use trusted recipients, protect config.json because it stores SMTP credentials, prefer HTTPS/authenticated Frigate access where available, and run it only in environments where people captured by the cameras have appropriate notice or consent.

SkillSpector

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (6)

Context-Inappropriate Capability

Medium

Confidence: 83% confidence
Finding: The skill executes external programs to forward detection data over WhatsApp and email, creating an outbound data-transfer channel from surveillance events to third-party services. In a home-security context this can expose sensitive presence, identity, and camera information if configured carelessly, compromised, or used without informed consent.

Context-Inappropriate Capability

Medium

Confidence: 88% confidence
Finding: The code fetches a snapshot over HTTP and writes it to a predictable local path before forwarding it by email. Because the image may contain sensitive surveillance data and the default API URL uses plain HTTP, the snapshot can be intercepted, tampered with, or exposed locally if host or network protections are weak.

Missing User Warnings

Medium

Confidence: 90% confidence
Finding: The skill explicitly handles surveillance events, face-recognition-derived identities, and image snapshots, then forwards that data to external WhatsApp and SMTP services, but the description does not clearly warn users about that sensitive data flow. This omission can lead to privacy, compliance, and consent failures because operators may deploy it without understanding that personally identifiable surveillance data leaves the local NVR environment.

Missing User Warnings

Medium

Confidence: 87% confidence
Finding: This script sends email and can attach an arbitrary local file path provided on the command line, but it gives no user-facing confirmation or warning that data is being transmitted to an external recipient. In an agent or automation context, this increases the risk of unintended data exfiltration, especially if upstream inputs can supply recipient addresses, message contents, or attachment paths without explicit user approval.

Missing User Warnings

Medium

Confidence: 90% confidence
Finding: The skill downloads camera snapshots and emails them externally without any visible consent, warning, or policy enforcement in the code. Surveillance images can contain highly sensitive personal data, so silent transmission increases privacy and compliance risk even if the functionality is intentional.

Missing User Warnings

Medium

Confidence: 86% confidence
Finding: The WhatsApp notification sends detection details such as identified person/object and camera name to an external messaging service without any user-facing disclosure in the code. Even text-only alerts can reveal occupancy patterns, identities, and surveillance coverage, making this a meaningful privacy exposure.

Static analysis

No suspicious patterns detected.