The Agent Skills Directory

[SAFE]: No malicious patterns or security vulnerabilities were identified in the skill files. The skill provides expert-level guidance for managing advertising campaigns and adheres to standard development practices.
[EXTERNAL_DOWNLOADS]: The skill includes instructions for users to clone its repository from GitHub (github.com/thatrebeccarae/claude-marketing.git). This is documented neutrally as a standard manual installation step.
[CREDENTIALS_UNSAFE]: The skill correctly handles sensitive configuration by providing a .env.example file with placeholders for API keys and tokens. It explicitly instructs users to manage these secrets via environment variables and warns against committing them to version control.
[PROMPT_INJECTION]: The 'Hard Rules' section in SKILL.md contains domain-specific constraints designed to ensure the quality of the advertising audit (e.g., requiring UET tag verification). These instructions do not attempt to bypass AI safety filters or override system-level directives.
[DATA_EXFILTRATION]: No evidence of unauthorized data exfiltration or suspicious network operations. The skill uses the official bingads Python library to communicate with Microsoft Advertising APIs.

microsoft-ads