Security audit

Adaptive Reasoning

Security checks across malware telemetry and agentic risk

Overview

This instruction-only skill openly adjusts the agent's reasoning behavior and output markers without requesting file, credential, network, shell, or persistent access.

Install this only if you want the agent to automatically spend more effort on complex prompts. Expect possible slower responses, higher token use, and reasoning icons at the end of some answers; avoid it if you want reasoning mode changed only by explicit user command.

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 (2)

Vague Triggers

High

Confidence: 97% confidence
Finding: The skill is configured to trigger on every user message, making it an always-on behavioral modifier rather than a narrowly scoped helper. In this context, that broad activation can silently alter response behavior across unrelated conversations, increase token consumption, and create a persistent avenue for instruction interference or policy drift.

Missing User Warnings

Medium

Confidence: 94% confidence
Finding: The skill directs the agent to enable reasoning automatically and explicitly says not to ask the user, which removes user awareness and control over a materially different processing mode. Because the skill also instructs use of internal commands or tools and appends covert state indicators, it can silently change runtime behavior in ways that are hard to audit and may conflict with higher-priority system behavior.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.