Protocol Deviation Classifier

Security checks across malware telemetry and agentic risk

Overview

This is a local clinical-trial deviation classifier with no evidence of data theft or hidden system access, but its high-stakes recommendations should be reviewed by qualified staff.

Install only if you want a local rule-based helper for protocol deviation triage. Do not let its classifications or recommended actions trigger sponsor, ethics committee, disciplinary, or investigator-blacklisting decisions without review under your organization’s SOPs, QA process, and legal or regulatory oversight. Avoid unnecessary pip installs from requirements.txt on modern Python, and use deidentified data whenever possible.

SkillSpector

By NVIDIA

Vulnerability Patterns

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

Findings (1)

Context-Inappropriate Capability

Medium

Confidence: 83% confidence
Finding: The skill goes beyond classification and emits prescriptive compliance actions, including severe recommendations like blacklisting researchers. In a clinical-trial context, users may over-trust automated output, leading to unsafe governance decisions, disproportionate disciplinary action, or improper regulatory escalation without human review.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal