Security audit

Daily Mood

Security checks for vulnerabilities and agentic risk

Overview

Daily Mood is a coherent local mood-message skill, but users should understand it stores mood labels and can be configured for recurring scheduled messages.

Install only if you are comfortable with local storage of each registered user's mood label and timestamps. Use explicit consent before registering users or adding cron jobs, review the broad trigger keywords if you run many skills, and delete or disable local profiles when users opt out.

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 (4)

Vague Triggers

Medium

Confidence: 94% confidence
Finding: The trigger list includes many generic phrases such as '早安', '晚安', 'encourage me', and 'send me a message' that are likely to appear in ordinary conversation and can cause unintended activation. In this skill, accidental activation is more concerning because the skill can register users for recurring messaging behavior and generate personalized emotional-support outputs, creating privacy and consent risks beyond a one-off response.

Vague Triggers

Medium

Confidence: 91% confidence
Finding: The usage section defines broad activation conditions like users saying they are tired, anxious, happy, or saying '早安 / 晚安', without clear boundaries distinguishing normal conversation from a deliberate request to invoke the skill. Because this skill supports scheduled multi-user pushes and profile-based personalization, ambiguous triggering can lead to unanticipated handling of emotional-state data and unwanted outbound messaging workflows.

Missing User Warnings

Medium

Confidence: 89% confidence
Finding: The skill advertises multi-user scheduled push notifications and storage of user language and mood profile data, but the description does not present a clear user-facing notice about persistent retention, automatic recurring outbound messages, consent, or how to opt out. In a context involving emotional-state tracking, this increases privacy and user-expectation risk because users may not realize their mood labels are stored and used for scheduled personalized messaging.

Vague Triggers

Medium

Confidence: 91% confidence
Finding: The manifest includes several broad, generic trigger keywords such as '寄语', '鼓励我', '陪伴', 'encourage me', and 'send me a message' that can overlap with ordinary conversational requests unrelated to this skill. In an agent environment, this increases the chance of unintended invocation or routing hijack, causing the skill to intercept benign user intent and respond in contexts where users did not explicitly ask for this package.

Static analysis

No suspicious patterns detected.