Security audit
Security checks for vulnerabilities and agentic risk
This skill coherently fetches public credit-card product data and recommends cards, with no evidence of hidden persistence, credential access, or destructive behavior.
Install only if you are comfortable with a Chinese-language credit-card recommendation flow that contacts CITIC's card data endpoint and returns bank application links with sid/platform parameters. Treat recommendations as product discovery, not financial advice, and verify annual fees, eligibility, rates, and application terms on the bank's official pages before applying.
The skill instructs the agent to use Bash/Python and consume external data, but the manifest declares no explicit tool scope or permission boundaries. That makes the skill over-privileged by default and increases the chance of unintended code execution or network access if the skill is invoked in the wrong context or later modified.
The trigger description includes broad everyday phrases like '办哪张卡' and similar generic recommendation requests, which can cause the skill to activate on loosely related conversations. Over-broad activation raises the risk of unintended tool use, unnecessary collection of user preference data, and execution of recommendation logic when the user did not explicitly request this skill.
The skill description and instructions assume Chinese-only interaction without checking the user's language preference. This can lead to user confusion, misinterpretation of financial recommendations, and accidental acceptance of actions or options the user does not fully understand.
The natural-language instructions, usage text, and examples are entirely in Chinese, which effectively forces a specific language for users interacting with the skill. There is no indication that the language is optional, user-selectable, or justified as a region-specific constraint.
This markdown file contains user-facing natural-language content exclusively in Chinese, and there is no indication that the skill is region-specific or that users can opt into this locale. Under the language/locale policy check, forcing a specific language without user choice can be a policy violation.
The script makes an outbound HTTP(S) request to a remote banking CDN/API to retrieve card data, but neither the CLI help text nor normal output clearly discloses that network access will occur. In an agent skill context, undisclosed external requests can surprise users, leak usage metadata such as IP/platform tags, and violate least-surprise or privacy expectations even if the destination appears legitimate.
No suspicious patterns detected.
