The Agent Skills Directory

[SAFE]: The skill operates purely as a set of natural language instructions for the agent to act as a copy editor. It does not perform any sensitive operations, command execution, or network calls.
[DATA_EXPOSURE]: The skill reads from '.agents/product-marketing-context.md' or '.claude/product-marketing-context.md' to gain context about the brand. This is a standard and safe practice for providing domain-specific context to an agent without exposing sensitive system information.
[INDIRECT_PROMPT_INJECTION]: While the skill processes user-provided copy (an external input source), it lacks the capabilities (such as shell execution or network access) that would make an indirect prompt injection attack viable. The output is limited to text rewriting and suggestions.

copy-editing