The Agent Skills Directory

[SAFE]: The skill focuses entirely on providing a software engineering workflow and coding guidelines. It does not perform network operations, file system modifications outside of standard coding tasks, or execute remote code.
[INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to read CONTEXT.md to align with the project's domain language. This is a standard practice for context-aware agents and represents a minimal attack surface where external project data could influence the agent's behavior. Boundary markers and human review of test names are recommended.

tdd