Review Code

Security checks across malware telemetry and agentic risk

Overview

This is an instruction-only code review helper with disclosed local notes and no evidence of hidden execution or data sending.

Install only if you want a review-focused agent behavior. If you enable its optional ~/review-code/ memory, approve only non-sensitive preferences and project constraints, and avoid saving secrets, tokens, private code, or confidential findings there.

SkillSpector

By NVIDIA

Vulnerability Patterns

Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
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)

Vague Triggers

Medium

Confidence: 91% confidence
Finding: The activation guidance uses broad phrases like PR checks, merge readiness, or bug-risk scans without a precise trigger boundary, which can cause the skill to activate in contexts the user did not explicitly request. In an agent setting, ambiguous auto-activation increases the chance of unsolicited behavior, context drift, or inappropriate persistence of review-mode assumptions across unrelated tasks.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal