The Agent Skills Directory

[SAFE]: No malicious behavior detected. The provided script scripts/refactor-scan.py performs deterministic static analysis on local files without any network or dynamic execution components.
[PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection as it ingests untrusted code from local source files and user-supplied arguments to generate refactoring plans and agent instructions. * Ingestion points: The scripts/refactor-scan.py script reads source files, and $ARGUMENTS in SKILL.md accepts user input. * Boundary markers: No specific delimiters or instructions to ignore embedded content are used in the prompt templates. * Capability inventory: The agent has access to Read, Edit, Grep, Glob, and Bash tools, and can create additional agent teams. * Sanitization: Content from the analyzed files is not escaped or sanitized before being processed by the agent.

refactor