knowledge-digest

Security checks across malware telemetry and agentic risk

Overview

This is an instruction-only study-material generator with no scripts, credentials, persistence, or hidden data handling, though users should be mindful that uploaded learning materials may be processed into generated files.

Install if you want a Chinese-language helper for generating study assets. Provide only learning materials you are comfortable having processed, and review generated quizzes, slides, notes, mind maps, and audio before sharing them.

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

Medium

Confidence: 90% confidence
Finding: The trigger list includes broad, everyday phrases such as '学习材料', '测验', and '幻灯片', which can cause the skill to activate in contexts where the user did not intend to invoke this specific capability. In an agent environment, accidental invocation can redirect user workflows, cause unintended file generation, or expose uploaded educational materials to the wrong skill pipeline.

Natural-Language Policy Violations

Medium

Confidence: 71% confidence
Finding: The skill is specified entirely in Chinese and does not define any language negotiation or fallback behavior, which can cause the agent to respond in an unexpected language or mis-handle multilingual user requests. This is primarily a safety and usability issue rather than a direct security flaw, but in practice it can increase the chance of user misunderstanding around generated outputs and consent for actions.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal