ClawHub
Back to skill

Security audit

master-ouyi

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed Buddhist teaching/persona guide with limited source lookup and no evidence of hidden, destructive, or credential-seeking behavior.

Install this if you want Ouyi-focused Buddhist teaching with strict citation behavior and occasional FoJin lookups. Be aware it may activate on related Tiantai/Pure Land topics and use a stylized Chinese Buddhist voice; it is not a substitute for personal religious instruction.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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 (3)

Vague Triggers

Medium
Confidence
93% confidence
Finding
The trigger logic is intentionally expansive: it says to invoke whenever a user's question merely 'touches' Ouyi-related themes, even without an explicit request. In an agent setting, that can cause over-activation, unwanted persona/style takeover, and irrelevant instruction injection into conversations that only loosely mention related concepts, reducing user control and increasing the chance the skill's stricter behavioral rules dominate unrelated tasks.

Natural-Language Policy Violations

Medium
Confidence
84% confidence
Finding
The skill hard-codes first-turn address/style constraints rooted in Chinese honorific conventions, without checking the user's language, locale, or stated preference. This can override user expectations, create confusing or culturally mismatched responses, and in multi-skill systems acts as an unnecessary behavioral constraint that may interfere with higher-priority user formatting or accessibility needs.

Natural-Language Policy Violations

Medium
Confidence
90% confidence
Finding
The skill hard-codes a single response style and persona for all matching queries, without checking whether the user wants that roleplay or linguistic register. This can override user preference, reduce transparency, and increase the chance of deceptive or manipulative responses when the assistant speaks in an authoritative religious voice by default.

VirusTotal

63/63 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.