Security audit
Security checks for vulnerabilities and agentic risk
The skill mainly searches public provider listings, but it asks for an API key and includes authenticated private-message guidance that goes beyond plain search.
Review this skill before installing if you only want public search. It requests MJZJ_API_KEY despite saying search does not need authentication, and its documentation can lead an agent from lookup into authenticated private messaging through another skill. Only provide the API key and allow contact actions when you intend the agent to message selected providers on your behalf.
The skill is advertised as a public service-provider search tool, but the documentation extends its workflow into authenticated private-message initiation via another skill. This broadens the effective capability scope from passive lookup to outbound contact, which can lead an agent to perform higher-risk actions than the user may expect from a search skill.
The documentation includes message-sending guidance that is not required for the stated purpose of searching providers. This creates capability creep and increases the chance that an agent transitions from information retrieval to action-taking without strong user intent verification.
The skill describes directly sending private messages to external parties without requiring an explicit warning or confirmation for an authenticated contact action. That can cause unauthorized or surprising outreach, especially when an agent chains the returned userSlug into a messaging API automatically.
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
curl -X GET "https://data.mjzj.com/api/spQuery/getClassifies" \
-H "Content-Type: application/json"
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
curl -X GET "https://data.mjzj.com/api/spQuery/getClassifies" \
-H "Content-Type: application/json"
The trigger and routing instructions are written entirely around Chinese phrases such as “卖家之家服务商”, “服务商查询”, and “查服务商”, and the description does not offer any language or locale choice. This can constitute a language/locale policy issue because the skill appears to hard-code Chinese-language invocation expectations rather than documenting an opt-in or region-specific constraint.
No suspicious patterns detected.