Security audit

auto-monitor

Security checks for vulnerabilities and agentic risk

Overview

This skill is a system-monitoring helper, but it asks for proactive monitoring and automatic repairs without clear user opt-in, scope, or rollback controls.

Install only if you are comfortable with an agent inspecting local or server health data and potentially acting on remediation guidance. Before use, define the monitored systems, check frequency, allowed commands, reporting destination, and require explicit approval before any repair action.

SkillSpector

Vulnerability Patterns

Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
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

Findings (4)

Vague Triggers

High

Confidence: 93% confidence
Finding: The trigger keywords are extremely broad, including common words like 'system', 'monitor', and 'auto', which can cause unintended invocation in unrelated conversations. In an agent environment, accidental activation of a proactive monitoring skill can lead to unnecessary system-state access, confusing behavior, or privacy-impacting status collection without clear user intent.

Vague Triggers

High

Confidence: 95% confidence
Finding: The skill explicitly says it should proactively monitor and report 'without waiting for instructions,' but it does not define strict activation conditions, schedule ownership, consent boundaries, or target scope. That ambiguity is dangerous because it normalizes autonomous system inspection and reporting in contexts where the user may not have requested persistent monitoring, increasing the chance of privacy, operational, or policy violations.

Missing User Warnings

High

Confidence: 90% confidence
Finding: The statement that issues can be 'automatically repaired immediately' encourages system-affecting actions without warning, confirmation, rollback guidance, or scope limits. Even though the current manifest is read-only, this instruction is hazardous as skill content because it promotes autonomous remediation behavior that could cause outages, destructive changes, or unauthorized modifications if execution capability is later added or inferred by a host agent.

Missing User Warnings

Medium

Confidence: 88% confidence
Finding: The skill promotes ongoing proactive monitoring but does not warn users that monitoring may collect operationally sensitive information such as process state, network status, or resource usage. Lack of transparency and consent can create privacy and operational risks, especially in shared or regulated environments.

Static analysis

No suspicious patterns detected.