Security audit

轻量级主动代理

Security checks for vulnerabilities and agentic risk

Overview

This skill is openly about background automation, but it gives broad unattended authority for commands, data checks, and external task actions without enough scoping or user-control detail.

Review carefully before installing. Only use it with least-privilege accounts, keep auto-claiming disabled unless you have explicit rules and approval gates, and do not allow custom commands, backups, mail access, or notifications to run unattended until you have audit logs, rollback, and a clear stop mechanism.

A.I.G

Vulnerability Patterns

Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs

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 (6)

Missing User Warnings

High

Confidence: 96% confidence
Finding: The skill describes automatic execution of scripts/commands, data syncing, backups, and periodic checks, but does not provide concrete warnings about the security and operational risks of unattended execution. This is dangerous because users may enable recurring tasks that access sensitive data or run destructive commands without understanding the need for least privilege, review, rollback, and monitoring.

Missing User Warnings

High

Confidence: 97% confidence
Finding: The skill explicitly supports automatically claiming and completing bounty tasks, which are external and potentially irreversible actions, but does not warn users about account, financial, reputational, or policy risks. Autonomous interaction with third-party systems can trigger unintended submissions or commitments at scale if misconfigured or abused.

Vague Triggers

Medium

Confidence: 94% confidence
Finding: The skill advertises broad autonomous background execution ('无需人工触发，自动在后台完成预设任务') without defining clear trigger boundaries, approval requirements, or scope limits. In an agent skill, this increases the risk of unattended actions being initiated in contexts the user did not fully anticipate, especially when combined with monitoring, command execution, and notifications.

Natural-Language Policy Violations

Medium

Confidence: 80% confidence
Finding: The file’s descriptive content is presented entirely in Chinese and does not indicate that users may choose another language or that the skill is intentionally limited to a Chinese-speaking or region-specific context. This can violate a language/locale choice policy when no opt-in or justification is provided.

Ae4

Medium

Category: analysis-evasion
Confidence: 80% confidence
Finding: Suspicious Unicode normalization or mixed-script content

Ae4

Medium

Category: analysis-evasion
Confidence: 80% confidence
Finding: Suspicious Unicode normalization or mixed-script content

Static analysis

No suspicious patterns detected.