The Agent Skills Directory

[SAFE]: The skill operates as a high-level advisory tool for marketing and product development. It contains no script execution, subprocess calls, or network connectivity.
[DATA_EXPOSURE]: The skill instructions specify reading project-related context files (e.g., .agents/product-marketing.md) if they exist. This is a standard pattern for context-aware AI agents and does not involve accessing sensitive system files or credentials.
[PROMPT_INJECTION]: No evidence of role-play overrides, safety bypasses, or system prompt extraction instructions was found. The instruction to respond in Russian is a localized behavior setting and not a malicious injection.

onboarding