The Agent Skills Directory

Indirect Prompt Injection Surface: The skill involves reading and verifying runtime log output via system tools such as log stream or Console.app. System logs can contain data from various external sources, which represents an ingestion point for untrusted content that could potentially influence agent behavior.
Ingestion points: Runtime logs accessed through the terminal or the Console application, as described in SKILL.md.
Boundary markers: The skill does not define explicit delimiters to separate log content from instructions, although it does recommend using specific predicates to filter for relevant subsystems and categories.
Capability inventory: The skill assumes the ability to execute shell commands to view logs and to modify application source code to add instrumentation.
Sanitization: The skill recommends using subsystem and category predicates to focus output, but does not specify sanitization of the log content itself.

telemetry