The Agent Skills Directory

[SAFE]: The skill's primary function is to transform text input into a structured Markdown document using local context templates. All operations are local to the ~/.claude/ directory.
[DATA_EXFILTRATION]: While the skill accesses ~/.claude/company-context.md, this is for the intended purpose of providing organizational context. No data is sent to external domains or third-party services.
[PROMPT_INJECTION]: The skill processes user-supplied topics or office-hours results. While these are external inputs, the skill's logic is restricted to document generation and does not expose high-privilege tool execution paths that could be manipulated by malicious input.

brief