InterCEPT SDR

• Outcome: Review
• Latest audit: May 28, 2026, 3:37 AM
• Version:
  1.0.0

Overview

This skill is a disclosed SDR control integration, but it grants broad live monitoring and process-control authority without enough scoping or safety guidance.

Install only if you intentionally want an agent to control your own iNTERCEPT SDR instance. Change default credentials, restrict the service to trusted networks, and require explicit user confirmation before scans, recordings, killall, remote-agent actions, or retrieval of decoded communications.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (5)

Vague Triggers

Medium

• Confidence: 91% confidence
• Finding: The skill description is extremely broad and includes many surveillance and control actions, increasing the chance the agent invokes it for loosely related SDR, radio, monitoring, or reconnaissance requests. Overbroad routing is dangerous here because the skill can start scanners, remote agents, recordings, and signal intelligence workflows with real-world monitoring implications.

Missing User Warnings

High

• Confidence: 95% confidence
• Finding: The documentation exposes destructive and surveillance-capable operations such as process termination, WiFi/Bluetooth scanning, TSCM sweeps, drone detection, remote agent control, and recordings without any safety constraints, confirmation requirements, or legality/authorization warnings. In this skill context, that is more dangerous because the API is expressly designed to operate SIGINT capabilities against live environments, so accidental or unauthorized use could disrupt systems or enable covert monitoring.

Natural-Language Policy Violations

Low

• Confidence: 98% confidence
• Finding: The skill instructs the agent to authenticate with hard-coded default credentials ('admin/admin'), effectively normalizing use of a shared default account. If the target instance still uses defaults, the skill enables trivial unauthorized access; if not, it encourages insecure deployment practices and unauthorized login attempts.

Missing User Warnings

Medium

• Confidence: 82% confidence
• Finding: The /adsb/aircraft endpoint is documented as a GET returning live aircraft position data without any stated authentication requirement, unlike nearby endpoints that explicitly require auth. In an SDR/SIGINT platform, exposing continuous location and movement data unauthenticated can enable unauthorized surveillance, aggregation, and tracking of nearby air traffic activity.

Missing User Warnings

Medium

• Confidence: 79% confidence
• Finding: The /acars/messages endpoint exposes decoded ACARS contents and is documented without any authentication requirement or warning, even though message bodies may contain operationally sensitive airline communications. In this skill context, decoded radio traffic is a core SIGINT output, so publishing message contents broadly increases privacy and intelligence-gathering risk.

VirusTotal

VirusTotal findings are pending for this skill version.