Security audit
Security checks for vulnerabilities and agentic risk
This looks like a WhatsApp formatting guide, but it also contains unrelated claims about file writing, API calls, and command execution that do not fit its stated purpose.
Install only if you want a Chinese WhatsApp formatting reference and are comfortable ignoring the unrelated automation, API, file-writing, and command-execution language. Do not let this skill authorize shell commands, network diagnostics, external API use, or file writes unless you separately asked for those actions and can review them.
The skill explicitly classifies itself as pure Markdown guidance with no execution requirements, yet later instructs the agent to run network diagnostic commands such as ping. This mismatch can mislead reviewers and users about operational behavior, causing an agent to attempt command execution outside the declared scope and tool permissions.
The skill says it does not involve WhatsApp/API calls, but other sections discuss API key handling, external service calls, and communication security as if they are part of normal operation. These contradictory claims can normalize broader data access and network behavior than users expect, increasing the chance of inappropriate secret handling or unreviewed integrations.
A WhatsApp formatting guide should not claim capabilities like file processing, API integration, and system command execution, especially when those capabilities are unrelated to the stated function. This kind of scope inflation is dangerous because it can disguise a broader-behavior skill as harmless documentation, undermining user consent and security review.
Unjustified references to file processing, API integration, and command execution materially expand the apparent privilege and behavior envelope of an otherwise simple text-formatting skill. In agent environments, that can prompt unnecessary access to local files, network resources, or shell features, creating avoidable attack surface.
The description says to use the skill for broad activities like '设计创作、UI设计、海报制作、品牌视觉', which are generic design requests rather than narrowly scoped WhatsApp formatting tasks. This makes invocation boundaries unclear and increases the risk that the skill is selected for unrelated work.
The description states '支持中文交互', and the document consistently frames the skill as Chinese-only without offering a language choice. Forcing a specific language or locale without user opt-in can violate organizational language policy requirements.
Marketing a simple guide as a high-efficiency automation workflow with batch file handling, API parsing, and multi-interface aggregation creates an inaccurate trust boundary around the skill. Even if no code is present, this misrepresentation can influence agent orchestration and user expectations toward broader, riskier behavior.
