The Agent Skills Directory

[SAFE]: The skill consists of instructional markdown files and reference playbooks for performance marketing. The content is strictly educational and instructional, providing templates and strategies for Google Ads, Meta, LinkedIn, and other platforms.
[CREDENTIALS_UNSAFE]: No hardcoded credentials were found. Code snippets for conversion tracking use clearly marked placeholders like 'YOUR_PIXEL_ID' or 'AW-XXXXXXXXX'.
[PROMPT_INJECTION]: Instructions include persona setting ('You are an expert performance marketer') and specific output constraints (RSA specifications), which are standard for guiding agent behavior and do not attempt to bypass safety filters or override core instructions.
[EXTERNAL_DOWNLOADS]: Reference files include standard tracking scripts for well-known services (Google, Meta, LinkedIn, TikTok). These are provided for documentation and user implementation, not for execution by the agent.
[INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to read product marketing context from local files. While this creates an ingestion surface, the skill does not possess dangerous capabilities (like arbitrary command execution or file writes) that would make this surface exploitable.

ads