The Agent Skills Directory

[SAFE]: The skill's primary function is to facilitate manual testing workflows. It uses a tool named mcp__gitnexus__query to analyze code and identifies relevant files for the testing process.
[DATA_EXPOSURE]: The skill writes test plans to the local file system under docs/testing-plan/. This is standard behavior for generating documentation and does not involve sensitive directories or exfiltration to external domains.
[PROMPT_INJECTION]: The skill contains strict process constraints (e.g., "Iron law: only execute one step at a time"), but these are operational guidelines for the AI agent rather than attempts to bypass system safety filters or override global instructions.

manual-testing