Security audit

Fun Skills

Security checks for vulnerabilities and agentic risk

Overview

This is mostly a local entertainment toolkit, but it claims scheduled retries to install other skills and includes file-overwriting setup examples without enough user control detail.

Review this before installing if you care about strict control over installed skills. The entertainment scripts themselves are low-risk, but run the initializer only if you are comfortable creating and overwriting files under ~/games, and ask the publisher to remove or clearly document the scheduled retry/install behavior for other skills.

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 (5)

Vague Triggers

Medium

Confidence: 92% confidence
Finding: The README suggests invoking the skill with very broad, everyday phrases like '我想玩个游戏' and '讲个笑话'. Such generic utterances are likely to overlap with normal conversation, increasing the chance of accidental activation and unintended execution of the skill or its connected behaviors. In this entertainment-skill context the impact is moderated because the described functions are not inherently privileged, but accidental triggering is still a genuine safety issue.

Missing User Warnings

Medium

Confidence: 89% confidence
Finding: The README instructs users to run commands that create local directories under ~/games without warning that this modifies the local filesystem. Even though mkdir -p is relatively low risk, it still changes user state and may be executed without informed consent; normalizing write operations in setup docs can lead to unexpected local data modification. The local-integration context makes this more relevant because the skill is explicitly designed to manage files on disk.

Missing User Warnings

Medium

Confidence: 97% confidence
Finding: The README includes a shell redirection example using '>' to write directly to ~/games/video/playing.md, which will overwrite the target file if it already exists. This is dangerous because users may copy-paste the command and unintentionally destroy existing notes or records; the risk is heightened by the absence of any overwrite warning or safer alternative. In a skill centered on maintaining personal game logs, silent data loss is especially relevant.

Vague Triggers

Medium

Confidence: 87% confidence
Finding: The manifest description uses very broad activation criteria such as 'user wants entertainment or fun activities' and 'looking for fun things to do,' which can cause the skill to be invoked in many loosely related contexts. Over-broad routing increases the chance of inappropriate activation, response hijacking of other more suitable skills, and unnecessary exposure of file-writing or command examples in contexts where they were not intended.

Natural-Language Policy Violations

Medium

Confidence: 79% confidence
Finding: The skill content is predominantly in Chinese and does not offer language negotiation or declare a locale restriction, which can lead to misapplication for users expecting another language. While this is not a direct security exploit, it can cause unsafe misunderstandings of instructions, file operations, or activity recommendations when the user's language context does not match.

Static analysis

No suspicious patterns detected.