Security audit

Asset Graph Management

Security checks for vulnerabilities and agentic risk

Overview

The skill is not malicious, but it encourages broad proactive infrastructure discovery from private entrypoints without enough scoping or confirmation.

Install only if you want an agent to help inventory infrastructure. Before using it with private systems, explicitly define the target assets, allowed credentials, discovery depth, prohibited data reads, where records may be written, and whether any generated catalog can be shared.

SkillSpector

Vulnerability Patterns

Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
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

Findings (2)

Vague Triggers

Medium

Confidence: 85% confidence
Finding: The skill description is extremely broad and is likely to activate for many ordinary infrastructure, operations, and discovery-related requests. That overbreadth can cause the agent to invoke a capability that encourages proactive environment discovery, increasing the chance of unnecessary access, data collection, or exposure of sensitive operational context beyond the user's intended scope.

Vague Triggers

Medium

Confidence: 91% confidence
Finding: The operating boundary says read-only discovery should proceed proactively after the user provides an asset, entrypoint, or access configuration, but it does not require a clear confirmation step or tightly scoped consent. In a security-sensitive infrastructure context, even read-only discovery can enumerate internal systems, collect metadata, traverse linked services, and expose confidential architecture details, making ambiguous auto-start behavior dangerous.

Static analysis

No suspicious patterns detected.