The Agent Skills Directory

[COMMAND_EXECUTION]: The validation script scripts/check-fixture.py uses subprocess.run to execute the Python unittest module. This is used exclusively to verify the logic of the TDD workflow against synthetic files in a temporary directory.
[DYNAMIC_EXECUTION]: The validation script scripts/check-fixture.py creates Python files from hardcoded strings and executes them at runtime. This dynamic behavior is contained within a test fixture and is used to illustrate the red-green-refactor cycle without external inputs.
[INDIRECT_PROMPT_INJECTION]: The skill processes external project code and instructions as part of its core TDD functionality.
Ingestion points: The skill instructions (SKILL.md) direct the agent to inspect target instructions, agreement, code, and test commands.
Boundary markers: The skill explicitly instructs the agent to resolve ambiguities or authority gaps with the technical lead rather than proceeding blindly.
Capability inventory: The agent is authorized to write implementation code and execute test suites, which are the primary capabilities of the skill.
Sanitization: There is no automated sanitization of ingested content, which is typical for development-focused skills that rely on human-guided task boundaries.

tdd