The Agent Skills Directory

[SAFE]: The skill defines a structured workflow for implementing tests, focusing on reliability and maintainability. It includes explicit rules against anti-patterns like testing internals or bundling unrelated assertions.
[SAFE]: No instances of prompt injection, data exfiltration, or obfuscation were found. The skill emphasizes security-friendly practices, such as prohibiting unsandboxed network access and unseeded randomness in tests to ensure determinism.
[SAFE]: Control flow is well-defined, and the skill instructs the agent to defer to the user for undefined commands rather than attempting to guess or execute unverified code. It also includes strong safety constraints, such as refusing to modify production code or issue review results on its own work.

write-testing