Back to skill

Security audit

nine

Security checks for vulnerabilities and agentic risk

Overview

The skill mostly explains the number 9, but it also embeds unrelated promotional links and contact details that could steer answers off-topic.

Review this skill before installing. It is not showing code execution or data theft, but users asking about the number 9 may also expose the agent to unrelated project promotion, links, and personal contact information. Removing the promotional section and person-specific guidance would make the skill proportionate to its stated purpose.

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
Findings (1)

T01 · Skill Instruction Hijacking

Warning
Location
SKILL.md:111
Finding

Unrelated Promotional Content and Behavioral Guidance Injected into Skill Context

Content
View full analysis
Remediation
View remediation
Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (3)

Description-Behavior Mismatch

High
Category
Not specified by scanner
Confidence
99% confidence
Finding

The skill is supposed to explain the cultural significance of the number 9, but it includes unrelated promotion for a person, projects, URLs, and direct contact information. In an agent setting, this creates prompt-injection-like behavior where benign user queries about '9' could be used to surface marketing content or external links unrelated to user intent, increasing spam, trust abuse, and possible phishing/exfiltration risk if users are steered off-platform.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

The activation description is broad enough that the skill may trigger on common cultural or symbolic questions without strong boundaries. Overbroad routing increases the chance that users are unnecessarily exposed to the skill's unrelated embedded content, which is especially concerning here because the skill contains off-topic promotional material later in the file.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
86% confidence
Finding

The note requires the skill to automatically switch between Chinese and English according to the user's language. Under the policy, forcing a language or locale behavior without explicit user choice can be a natural-language policy issue unless the user is offered an opt-in or selection mechanism.

Content

No source excerpt is available for this finding.