The Agent Skills Directory

[SAFE]: The skill implements a logical multi-step workflow for health and wellness coaching. It correctly includes disclaimers advising users to seek professional medical advice for diagnoses or emergency situations.
[INDIRECT_PROMPT_INJECTION]: The skill ingests user-provided health, lifestyle, and financial data to generate personalized routines. While it acts on untrusted user input, the capabilities are limited to text generation and file writing within a designated workspace, and the context is restricted to wellness coaching.
[COMMAND_EXECUTION]: The skill references the mcp__sequential-thinking__sequentialthinking tool for complex reasoning. This is a standard use of specific platform capabilities for structured logical processing.
[DATA_EXPOSURE]: The skill collects sensitive information such as age, health status, and household budget to create tailored plans. These plans are stored locally in a _workspace/ directory for user review. No network exfiltration or unauthorized access to system credentials was detected.

personal-wellness