The Agent Skills Directory

[PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection via the following: 1. Ingestion points: The skill fetches untrusted content from social media (X, Reddit, Hacker News), competitor websites, and various external feedback sources in loops like newsjacking, social-listening, and competitor-watch as defined in references/loop-catalog.md. 2. Boundary markers: No explicit instructions for delimiters or 'ignore instructions' guards are provided for the processed data. 3. Capability inventory: The skill orchestrates high-privilege capabilities including budget management (ads skill), outbound communication (emails skill), and public posting (social skill). 4. Sanitization: There is no requirement for sanitizing or escaping ingested data before it is interpolated into prompts.
[COMMAND_EXECUTION]: The skill instructs the agent to use platform-native scheduling commands such as /loop, ScheduleWakeup, and CronCreate (for Claude Code) to operationalize the marketing workflows. This involves the programmed execution of system-level primitives to manage the agent's task lifecycle.

marketing-loops