The Agent Skills Directory

[SAFE]: The skill strictly adheres to its stated purpose of providing refactoring guidance and patterns. All provided examples are consistent with standard software engineering best practices.
[COMMAND_EXECUTION]: The documentation includes standard development commands such as npm test and git commit. These are contextually appropriate for the refactoring workflows described and do not represent a security risk.
[INDIRECT_PROMPT_INJECTION]: The skill outlines a conceptual architecture for ingesting and acting upon data from other agents (e.g., qe-quality-analyzer). While this creates a surface for processing external tool outputs, it is presented as a standard coordination pattern within the fleet manager context.
Ingestion points: Pseudo-code in SKILL.md describes ingesting output from qe-quality-analyzer.
Boundary markers: None explicitly defined in the documentation examples.
Capability inventory: Mentions Task execution and FleetManager coordination.
Sanitization: None explicitly shown in the conceptual integration examples.

qe-refactoring-patterns