Multi Agent Pipeline

Security checks across malware telemetry and agentic risk

Overview

This skill is advertised as a generic no-network pipeline, but the included code is a specialized voice and story API that sends audio/text to external providers and stores user content.

Install only if you intend to use the bundled story and voice API, not just a generic pipeline template. Review the code, provider terms, API key handling, database/cache behavior, and privacy implications for audio and child-related content before enabling it.

SkillSpector

By NVIDIA

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (2)

Tp4

High

Category: MCP Tool Poisoning
Confidence: 87% confidence
Finding: A substantial description-behavior mismatch is a real security concern because it can hide materially different functionality than users or reviewers expect, including provider-specific integrations, file/audio handling, storage, and database-backed endpoints. Mislabeling a specialized, networked, media-processing workflow as a generic provider-agnostic pipeline can cause operators to grant trust, permissions, or deployment approval under false assumptions.

Missing User Warnings

Medium

Confidence: 85% confidence
Finding: Uploaded voice audio is sent to external STT providers, which may include sensitive personal or biometric data. In this skill context, the service appears consumer-facing and child-focused, so undisclosed third-party transmission increases privacy and compliance risk, especially if users do not explicitly consent to cloud processing.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal