The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses a Python shell command to validate the YAML frontmatter of the target documentation file. While it uses yaml.safe_load(), which is a recommended security practice to prevent arbitrary code execution during parsing, any shell execution involving external file paths carries a theoretical risk of command injection depending on how the execution environment handles the file path argument.
[PROMPT_INJECTION]: The skill processes untrusted data from existing documentation files. This presents an indirect prompt injection surface where malicious instructions embedded in a document (e.g., hidden in comments or prose) could attempt to influence the sub-agents or the main agent's behavior during the refactoring process.
Ingestion points: Reads the file content from the path provided in $ARGUMENTS or user input.
Boundary markers: None (the skill content is read and passed directly to analysis agents without explicit delimiters or 'ignore' warnings).
Capability inventory: The agent has the capability to read files, write/edit files, and execute shell commands via Python.
Sanitization: There is no explicit sanitization or filtering of the document content before it is processed by the AI agents.

refactor-doc