The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze external data in the form of user-provided writing samples. This creates a surface for indirect prompt injection where instructions hidden within samples could attempt to influence the agent's analysis. However, the skill includes specific formatting rules, such as wrapping samples in code blocks, which helps mitigate accidental instruction following. As this data ingestion is the primary purpose of the skill, the risk is considered low and manageable within the context of profile creation.
[COMMAND_EXECUTION]: The skill instructions direct the agent to write and update Markdown files in the local project directory (e.g., about-me.md, voice.md, and files within the profiles/ directory). These file system operations are localized, transparent to the user, and strictly aligned with the documented purpose of persisting the generated voice profiles.

skill-voice-builder