The Agent Skills Directory

[SAFE]: No malicious patterns or security risks were identified. The skill follows standard engineering practices for testing and refactoring.
[COMMAND_EXECUTION]: The skill utilizes uv to run development tools such as pytest, mypy, and ruff. It also uses git commands for version control throughout the refactoring loop.
[EXTERNAL_DOWNLOADS]: The skill references standard Python development libraries including pytest, mypy, ruff, and approvaltests from well-known package registries.
[PROMPT_INJECTION]: The skill contains an inherent surface for indirect prompt injection as it processes local source code and test data to generate characterization tests.
Ingestion points: Reads local source code and user-provided data in SKILL.md and templates/characterization_tests.py.
Boundary markers: No explicit boundary markers or instruction-guardrails are defined in the code templates.
Capability inventory: Executes shell commands via uv and git; writes characterization files (golden masters) to the local filesystem.
Sanitization: No explicit sanitization or validation of the ingested code or data is performed.

test-driven-refactoring