The Agent Skills Directory

[SAFE]: The skill instructions and metadata do not contain any malicious patterns such as prompt injection, obfuscation, or unauthorized data access. The content is focused on providing architectural and logical guidance for software testing.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to process and modify repository content (test files and source code), which represents an inherent surface for indirect prompt injection. However, this is part of the skill's primary function for code refactoring, and it does not involve high-risk operations like network exfiltration or unauthorized file access. The instructions provide logical constraints for test generation which help mitigate accidental behavior changes.

test-hygiene