The Agent Skills Directory

[SAFE]: The skill establishes a functional workflow for managing reminders using the qqbot_remind and cron tools. This behavior aligns perfectly with the described purpose of the skill.
[SAFE]: Analysis of identifiers and data flow shows that sensitive information is handled correctly. Platform identifiers like openid are used solely for routing messages within the QQBot environment.
[SAFE]: An indirect prompt injection surface exists because user-supplied reminder content is stored and later used in an agent prompt template. However, this is inherent to the functionality of a reminder service and is managed using isolated session targets. * Ingestion points: The content parameter in qqbot_remind and the {提醒内容} placeholder in the cron payload. * Boundary markers: No explicit separators are used to isolate user content from the instructional template. * Capability inventory: The skill relies on the cron tool for persistence and the qqbot_remind tool for task configuration. * Sanitization: User input is not explicitly sanitized before being included in the scheduled task's message payload.

qqbot-remind