The Agent Skills Directory

[SAFE] (SAFE): The skill file is composed entirely of markdown documentation and frontmatter metadata. No malicious patterns, encoded content, or suspicious logic were detected.
[NO_CODE] (INFO): No executable scripts, shell commands, or programming logic are contained within this file. It is a pure instructional resource.
[PROMPT_INJECTION] (SAFE): The instructions are standard technical guidelines. There are no attempts to override agent constraints, ignore previous instructions, or escalate privileges.
[DATA_EXFILTRATION] (SAFE): No network requests, hardcoded credentials, or sensitive file path references were identified.
[INDIRECT_PROMPT_INJECTION] (LOW): The skill is intended to be used while reviewing or writing tests (external data). However, it lacks executable capabilities (write ops, network calls) that would make it a high-risk surface for indirect injection.

playwright