ClawHub
Back to skill

Security audit

model-pyramid

Security checks for vulnerabilities and agentic risk

Overview

This skill is an advisory model-and-effort sizing guide with a local validation script, and its sensitive cost/context implications are disclosed and user-controlled.

Before installing, treat this as guidance for model, effort, and advisor choices, not an automatic controller. Pay attention to the disclosed cost and context effects: advisors may receive the full conversation, higher effort can increase tool use and token spend, and some runtime settings may persist if you apply them yourself.

SkillSpector

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
91% confidence
Finding
The README presents the skill as bilingual, but the substantive operational guidance is predominantly in Chinese without an explicit language-selection mechanism. This can cause an agent or user to miss important behavioral constraints or safety-relevant details, leading to incorrect application of the skill and degraded review quality.

Static analysis

No suspicious patterns detected.