Security audit
Security checks for vulnerabilities and agentic risk
This prompt-only skill helps organize package deliveries using user-provided details and does not include code, network access, credential use, or persistence.
Installers should understand that the skill may help draft messages containing package, address-type, name/unit, schedule, or helper details. Review any drafted message before sending and avoid sharing pickup codes, account links, or unnecessary personal information.
The verification rule prohibiting any CJK characters is an unjustified language/script restriction that can exclude legitimate user content, names, addresses, or delivery notes. In a package-coordination context, this can cause denial of service, mishandling of multilingual delivery information, and discriminatory behavior without any security basis.
The skill repeatedly defines itself as a prompt-only workflow that must not access external systems, but later includes an exception allowing message sending through an approved channel. That contradiction can cause an agent to overstep its intended boundaries and initiate third-party communications based on ambiguous authorization, creating a social-engineering and data-disclosure risk.
The skill states it will not contact carriers, neighbors, building staff, sellers, or anyone else, yet later contemplates sending messages to such parties. This inconsistency can be misread by an autonomous agent as permission to contact third parties, potentially exposing delivery details, names, unit numbers, schedules, or access-related information without a clear safety gate.
The file declares "Language: en", which is a natural-language locale constraint. There is no indication that users may choose another language or that English-only behavior is required for a documented regional or compliance reason.
The manifest declares language: "en", which is a natural-language locale constraint. The file does not state that other languages are supported, offer user choice, or justify why English-only behavior is required for this skill.
No suspicious patterns detected.
