The Agent Skills Directory

[SAFE]: The skill establishes a methodology for Test-Driven Development (TDD) using the Red-Green-Refactor cycle. It is designed to be framework-neutral, delegating specific test execution commands to specialized environment tools rather than implementing its own execution logic.
[INDIRECT_PROMPT_INJECTION]: The skill contains a defensive Security Model section that addresses the risks of processing untrusted content. * Ingestion points: The skill processes existing production code, test suites, fixture data, snapshots, and test runner outputs as described in SKILL.md. * Boundary markers: The Security Model provides explicit instructions to distinguish between instructional requirements and data processed during the TDD cycle. * Capability inventory: Capabilities are limited to running targeted test commands and performing temporary code mutations for sensitivity testing. * Sanitization: The instructions require the agent to ignore directives embedded within test failures or code comments and to stick strictly to user-defined scopes.

tdd