unit-tests
Pass
Audited by Gen Agent Trust Hub on Oct 27, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill serves as a technical style guide and documentation for software testing practices. It does not contain malicious code or instructions intended to compromise the agent or the host environment.
- [DATA_EXFILTRATION]: The skill actively promotes data security by instructing the agent to use isolated environment variables (such as HOME, XDG, and TMPDIR) and to keep Git identities localized to specific repositories. This prevents the accidental exposure of global credentials or sensitive system information during the testing process.
- [INDIRECT_PROMPT_INJECTION]: While the skill involves processing and writing code files, it outlines robust isolation and sanitization techniques, such as using isolated environments for spawned processes, which reduces the risk of malicious input influencing the agent's behavior.
- [COMMAND_EXECUTION]: References to shell commands (e.g.,
bun test,git init,turbo) are constrained to standard developer workflows for project verification and automation within the workspace context.
Audit Metadata