The Agent Skills Directory

[SAFE]: The skill is entirely composed of documentation and guidelines for TDD. Its purpose is to provide the agent with a reference for writing better tests and following the 'red-green-refactor' cycle.
[INDIRECT_PROMPT_INJECTION]: The skill suggests reading project-specific files such as GLOSSARY.md and Architecture Decision Records (ADRs). This is a standard practice for maintaining domain consistency within a codebase. There are no patterns indicating that the agent would execute instructions found within these files as commands, and the skill itself does not expose any dangerous capabilities.
[DATA_EXPOSURE_AND_EXFILTRATION]: Code examples in mocking.md and tests.md use generic placeholder functions like fetch() and environment variables like STRIPE_KEY for educational purposes. These are used to illustrate how to structure code for testability and do not represent hardcoded secrets or attempts to exfiltrate data.

tdd