Security audit
Security checks for vulnerabilities and agentic risk
The skill does what it claims for 卖家之家 Q&A, but its broad trigger rules can cause unintended posting, account lookup, or token-backed API use.
Review before installing. Only use this skill if you intend your agent to access 卖家之家 with MJZJ_API_KEY. Be especially careful with generic requests like asking a question or searching questions, because the skill may route them to this service; confirm before posting questions or using bounty/watch payment fields.
The trigger guidance includes broad natural-language phrases such as “提问 / 搜索问题 / 我的问答”, which can match common user intents outside this specific service. In an agent environment, that can cause the skill to activate unexpectedly and route user data or actions to this external API when the user did not clearly intend to use 卖家之家.
The phrase-to-endpoint mapping lacks clear disambiguation boundaries, so generic requests like “查找问题” or “帮我提问” may invoke this skill even when the user means a different site, system, or internal task. Because some mapped actions include account-specific retrieval and question posting, ambiguous activation increases the risk of unintended data access or unintended external side effects.
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/ask/getCategories" \
-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/ask/getCategories" \
-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/ask/getCategories" \
-H "Content-Type: application/json"
This endpoint transmits user-authored content and a bearer token to an external service and performs a state-changing action (posting a question). While that is core functionality, it becomes security-relevant because broad trigger rules elsewhere in the skill could cause unintended posting or unintended disclosure of user content to the third party.
若发布失败，提示用户：
发布未成功，你可以前往 https://mjzj.com/ask/create 手动发布。
curl -X POST "https://data.mjzj.com/api/ask/createQuestion" \
-H "Authorization: Bearer $MJZJ_API_KEY" \
-H "Content-Type: application/json" \
-d '{
This endpoint sends an authorization token to retrieve the user’s private posted questions from a third-party service. Access to private/account-scoped data is expected for functionality, but in combination with the skill’s broad activation rules it creates a real risk of unintended private-data retrieval from the external platform.
倒序游标分页。首次请求
position传空字符串或不传；后续传上一页返回的nextPosition。
curl -X GET "https://data.mjzj.com/api/ask/queryMyQuestions?size=20&position=" \
-H "Authorization: Bearer $MJZJ_API_KEY" \
-H "Content-Type: application/json"
No suspicious patterns detected.