Paul Graham 视角

Security checks across malware telemetry and agentic risk

Overview

This is a low-authority advisory skill that provides a Paul Graham-inspired perspective using bundled text references, with no executable behavior or sensitive access.

Install this if you want a Chinese-first PG-inspired advisory lens for startup, writing, product, or founder questions. Be aware it defaults to Chinese unless overridden, and verify any time-sensitive market or startup facts separately.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • 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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (1)

Natural-Language Policy Violations

Medium
Confidence
93% confidence
Finding
The skill explicitly sets Chinese as the default output language unless the user specifies otherwise. This can override or conflict with user expectations and upstream system defaults, creating a prompt-safety and usability issue where the assistant may respond in an unintended language. In this skill’s context, the risk is limited because it affects presentation rather than granting capabilities or causing code execution, but it is still a real policy-alignment problem.

VirusTotal

59/59 vendors flagged this skill as clean.

View on VirusTotal