The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill workflows involve reading and displaying message history and active alerts, which are ingestion points for potentially untrusted data.
Ingestion points: The homelab_ntfyTopics tool is used to list recent messages across topics, and homelab_alertList identifies active alerts which may include user-controlled or externally-influenced strings (SKILL.md).
Boundary markers: There are no explicit delimiters or instructions provided to the agent to prevent it from obeying commands that might be embedded within a notification message or an alert description.
Capability inventory: The skill allows the agent to send further notifications via homelab_ntfySend and interfaces with other system-level skills like service-monitoring and log-analysis (SKILL.md).
Sanitization: The workflow does not mention sanitizing or escaping the content retrieved from homelab_ntfyTopics before the agent processes it.
[SAFE]: The skill uses clear, standard tools and practices for homelab management. References to external documentation (ntfy.sh) target well-known, legitimate services and do not represent a security risk.

notification-workflows