ACMG Variant Classifier

• Outcome: Pass
• Latest audit: May 29, 2026 at 8:32 PM UTC
• Version:
  1.0.0

Overview

This is a local ACMG variant-classification workflow with sensitive clinical inputs but no evidence of hidden data access, exfiltration, or unsafe automatic behavior.

Use this only as review-oriented decision support for germline SNV/indel interpretation. Avoid unnecessary patient identifiers, keep any generated records in approved clinical storage, and validate criteria against current ACMG/AMP and ClinGen guidance before relying on outputs.

SkillSpector

By NVIDIA

Vulnerability Patterns

• MCP Tool Poisoning: Hidden Instructions, Unicode Deception, Parameter Description Injection
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (1)

Tp4

High

• Category: MCP Tool Poisoning
• Confidence: 84% confidence
• Finding: The skill claims to provide a guided ACMG/AMP interpretation workflow, but the underlying behavior reportedly only consumes pre-aggregated evidence counts and outputs a label, without collecting evidence, assigning criteria from raw facts, or producing the promised review-ready summary. In a clinical decision-support setting, this mismatch can mislead users into overtrusting incomplete automation, causing incorrect variant classification or omission of important caveats and conflict resolution.

VirusTotal

65/65 vendors flagged this skill as clean.