The Agent Skills Directory

Prompt Injection (SAFE): No instructions attempting to override agent behavior or bypass safety filters were detected. The content focuses strictly on professional CRM and automation architecture.
Data Exposure & Exfiltration (SAFE): No hardcoded credentials, sensitive file paths, or network operations (curl, wget, etc.) are present. The mention of tokens like {first_name} refers to standard email marketing placeholders, not sensitive system data.
Obfuscation (SAFE): The text is clear, human-readable markdown with no evidence of Base64 encoding, zero-width characters, or homoglyph attacks.
Unverifiable Dependencies & Remote Code Execution (SAFE): No package managers (npm, pip) or remote script execution patterns are used. The skill contains no executable scripts.
Indirect Prompt Injection (SAFE): While the skill is designed to process user input related to marketing, it lacks executable capabilities (like file writing or network calls) that could be exploited via malicious data ingestion.

crm-automation