ClawHub
Back to skill

Security audit

AI Mood Journal - 情绪日记

Security checks for vulnerabilities and agentic risk

Overview

This mood-journal skill is not malicious, but it asks the agent to collect and maintain sensitive emotional and mental-health profiles without clear consent, limits, or safety boundaries.

Review carefully before installing. Use it only if you are comfortable having an AI assistant collect and summarize emotional, trigger, body-sensation, and coping-pattern information over time. It should not be treated as professional mental-health care, and users should avoid storing sensitive details unless the platform provides clear privacy, retention, and deletion controls.

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
  • 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 (3)

Missing User Warnings

High
Confidence
95% confidence
Finding
The skill instructs the agent to maintain a long-term emotional profile containing highly sensitive mental-health data, triggers, coping behaviors, and inferred patterns, but provides no warning, consent flow, retention limits, or privacy safeguards. This is dangerous because the skill encourages persistent profiling of special-category personal data, increasing the risk of privacy harm, misuse, overcollection, and unauthorized downstream use of intimate psychological information.

Natural-Language Policy Violations

Medium
Confidence
91% confidence
Finding
The skill is framed as '你的AI心理咨询师' and a 'personal psychotherapist,' which overstates capability and implies professional mental-health support without qualification, scope limits, or user opt-in. In the context of emotional distress, this can create unsafe reliance, discourage users from seeking real care, and lend undue authority to speculative psychological interpretations like identifying fears or behavioral patterns.

Vague Triggers

Medium
Confidence
95% confidence
Finding
The activation condition is very broad ('when the user wants to record emotions, analyze mental state, or track mood changes'), which can cause the skill to engage in sensitive mental-health contexts without clear boundaries or explicit consent. In a mental-health skill, over-triggering is riskier because it may collect highly sensitive psychological data or present quasi-therapeutic guidance in situations where the user did not intend to enter a journaling workflow.

Static analysis

No suspicious patterns detected.