ClawHub
Back to skill

Security audit

StrategyLens

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed quantitative-finance research and backtesting helper, with no hidden trading, credential, persistence, or destructive behavior found.

Install this for research-style quantitative strategy explanation and backtest templates, not for personalized investment advice or live trading. Users should verify any MCP data source they configure and keep broker/order-entry tools separate, since the skill itself is intended only for historical analysis and education.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (3)

Vague Triggers

Medium
Confidence
78% confidence
Finding
Overly broad trigger phrases increase the chance that the skill activates during generic finance discussions, causing unintended invocation of finance-specific guidance and MCP-backed market-data workflows. In an agent environment, broad activation widens the attack surface for prompt steering, context hijacking, or accidental use outside the user's actual intent.

Vague Triggers

Medium
Confidence
82% confidence
Finding
Ambiguous activation guidance without clear boundaries can cause the skill to run in contexts where it should not, including broad finance conversations or requests that drift toward advice-like outputs. Because this skill can call an MCP for historical market data and produce executable backtest templates, unclear activation criteria make accidental or adversarial invocation more likely.

Natural-Language Policy Violations

Medium
Confidence
89% confidence
Finding
The file explicitly instructs the skill to produce Chinese-language methodology explanations, which can override or constrain a user's requested language. This is a real policy/UX safety issue because hard-coded output-language requirements reduce user control and may cause the agent to ignore user preferences, though it does not directly create code-execution or data-exfiltration risk.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.