Security audit

triz-innovation-pro

Security checks for vulnerabilities and agentic risk

Overview

This skill is a coherent TRIZ analysis workflow that uses named analysis tools for its stated purpose and does not show hidden execution, persistence, or destructive behavior.

Install only if you are comfortable having technical problem descriptions, generated analysis, and selected solution details processed by the TRIZ analysis tools. Do not include trade secrets, regulated data, or sensitive proprietary design information unless your organization allows that use.

SkillSpector

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 (2)

Missing User Warnings

Medium

Confidence: 95% confidence
Finding: The skill explicitly directs the agent to send aggregated problem summaries, core problem descriptions, functional model data, user modification requests, and existing analysis data to the `TrizAnalysis` tool without any instruction to minimize, redact, classify, or disclose that transfer to the user. In an enterprise or proprietary engineering context, these fields can contain sensitive design details, internal failure analyses, or confidential user-provided content, so silent forwarding creates a real data exposure risk.

Vague Triggers

Medium

Confidence: 90% confidence
Finding: The refresh intent examples are broad enough to overlap with ordinary conversational follow-ups, which can cause the agent to enter the refresh branch when the user is merely asking for clarification or additional detail. In this skill, that can lead to unintended stateful behavior such as excluding previously shown solutions, applying preference filters, or presenting different recommendations than the user actually requested.

Static analysis

Detected: suspicious.prompt_injection_instructions

Prompt-injection style instruction pattern detected.

Warn

Code: suspicious.prompt_injection_instructions
Location: SKILL.md:114