Security audit
Security checks for vulnerabilities and agentic risk
This is a Xianyu listing-content helper with misleading auto-publishing wording, but it ships only instructions and requests no account, tool, or persistence access.
Installers should treat this as a drafting assistant, not an automatic Xianyu posting tool. Review generated listings before posting manually, and redact personal information, order numbers, addresses, serial numbers, IMEI numbers, and account identifiers from any proof-of-purchase images.
The skill metadata and headline advertise automatic publishing to Xianyu, while the FAQ states publishing is manual only due to platform restrictions. This mismatch can mislead users and higher-level agents into granting broader permissions or assuming account-posting actions will occur, creating unsafe expectations around marketplace automation.
The document contradicts itself about whether automatic publishing is supported, which undermines safe agent behavior and informed user consent. In an ecommerce/account context, contradictory instructions can cause an orchestrator or user to overtrust the skill with account actions, credentials, or posting workflows it should not perform.
The skill promotes listing publication and recommends sharing purchase-proof imagery, but it does not warn about privacy, account risk, or sensitive data exposure. Receipts, invoices, and purchase proofs may contain names, addresses, order numbers, IMEI/serial numbers, or other identifiers that could be misused for fraud or account compromise.
The trigger examples are broad and do not clearly define when the skill should activate versus when ordinary conversation should continue. In practice, this can cause over-invocation on ecommerce-related prompts and lead the system to solicit or process unnecessary product, account, or transaction details without clear user intent.
The manifest description is written only in Chinese ("闲鱼自动上架助手"), which may impose a language-specific experience without offering any user choice or documenting a justified locale constraint. The policy requires flagging language or locale constraints when they are forced without opt-in.
No suspicious patterns detected.
