The Agent Skills Directory

[SAFE]: The skill performs legitimate project planning and documentation tasks. It creates and manages files within a restricted docs/plans/ directory and does not modify source code or configuration files.
[COMMAND_EXECUTION]: System commands are used exclusively for generating timestamps to maintain organized, chronologically versioned plan artifacts.
[PROMPT_INJECTION]: The skill processes project content and user requirements. While this represents an indirect prompt ingestion surface (ingesting code in Step 1 and Step 5), the risk is minimal as the output is restricted to human-readable Markdown documentation. The skill does not contain sanitization or boundary markers for the processed project data, but its capabilities are limited to writing documentation artifacts.

write-plan