The Agent Skills Directory

[SAFE]: The skill consists of instructional templates and guidelines for digital marketing. No malicious behavior or security risks were identified.
[PROMPT_INJECTION]: No attempts to override system prompts, bypass safety filters, or extract sensitive instructions were found in any of the skill files.
[DATA_EXFILTRATION]: No hardcoded credentials, sensitive environment variables, or unauthorized network exfiltration patterns were detected. References to marketing providers (e.g., Google Ads, Meta) are for legitimate instructional purposes.
[REMOTE_CODE_EXECUTION]: The skill does not download or execute remote scripts. The JavaScript files in the tests/ directory are local validation utilities that perform basic file existence checks and do not interact with the network or external dependencies.
[COMMAND_EXECUTION]: There are no high-risk command execution patterns, privilege escalation attempts (sudo), or persistence mechanisms (cron jobs) found in the skill package.
[OBFUSCATION]: No obfuscated content, hidden Unicode characters, or encoded strings (such as Base64-encoded commands) were discovered during the analysis.
[INDIRECT_PROMPT_INJECTION]: While the skill defines ingestion points for external marketing data (objectives, budgets, landing pages), it does not possess high-risk capabilities that would allow this data to trigger dangerous system operations. The risk is considered negligible.

paid-media